Title of Invention	"FUSION PROTEINS COMPRISING A NON-CYTOTOXIC PROTEASE, A TARGETING MOIETY, A PROTEASE CLEAVAGE SITE AND A TRANSLOCATION DOMAIN"
Abstract	The invention provides a single chain, polypeptide fusion protein, comprising: a non-cylotoxic prolease, or a fragment threrof, which protease or protease fragment is capable of cleaving a protein of the exocytic fusion apparatus of a target cell; a Targeting Moiety that is capable odf binding to a Binding Sire on the target cell, which Binding Site is capable of undergoing endocytosis to be ncorporated into an endosome within the target cell; a protease cleavage site at which site the fusion protein is cleavable by aprotease, wherein the protease cleavage site located between the non-cytotoxic protease or fragment thereof and the Targeting Moiety; and a translocation domain that is capable of translocating the protease or protease fragment from within an endosome, across the endosomal membrane and into the cytosol of the target cell.

Title of Invention

"FUSION PROTEINS COMPRISING A NON-CYTOTOXIC PROTEASE, A TARGETING MOIETY, A PROTEASE CLEAVAGE SITE AND A TRANSLOCATION DOMAIN"

Abstract

The invention provides a single chain, polypeptide fusion protein, comprising: a non-cylotoxic prolease, or a fragment threrof, which protease or protease fragment is capable of cleaving a protein of the exocytic fusion apparatus of a target cell; a Targeting Moiety that is capable odf binding to a Binding Sire on the target cell, which Binding Site is capable of undergoing endocytosis to be ncorporated into an endosome within the target cell; a protease cleavage site at which site the fusion protein is cleavable by aprotease, wherein the protease cleavage site located between the non-cytotoxic protease or fragment thereof and the Targeting Moiety; and a translocation domain that is capable of translocating the protease or protease fragment from within an endosome, across the endosomal membrane and into the cytosol of the target cell.

Full Text	Fusion Proteins This invention relates to non-cytotoxic fusion proteins, and to the therapeutic application thereof. Toxins may be generally divided into two groups according to the type of effect that they have on a target cell. In more detail, the first group of toxins kill their natural target cells, and are therefore known as cytotoxic toxin molecules. This group of toxins is exemplified inter alia by plant toxins such as ricin, and abrin, and by bacterial toxins such as diphtheria toxin, and Pseudomonas exotoxin A. Cytotoxic toxins have attracted much interest in the design of "magic bullets" (eg. immunoconjugates, which comprise a cytotoxic toxin component and an antibody that binds to a specific marker on a target cell) for the treatment of cellular disorders and conditions such as cancer. Cytotoxic toxins typically kill their target cells by inhibiting the cellular process of protein synthesis. The second group of toxins, which are known as non-cytotoxic toxins, do not (as their name confirms) kill their natural target cells. Non-cytotoxic toxins have attracted much less commercial interest than have their cytotoxic counterparts, and exert their effects on a target cell by inhibiting cellular processes other than protein synthesis. Non-cytotoxic toxins are produced by a variety of plants, and by a variety of microorganisms such as Clostridium sp. and Neisseria sp. Clostridial neurotoxins are proteins that typically have a molecular mass of the order of 150 kDa. They are produced by various species of bacteria, especially of the genus Clostridium, most importantly C. tetani and several strains of C. botulinum, C. butyricum and C. argentinense. There are at present eight different classes of the clostridial neurotoxin, namely: tetanus toxin, and botulinum neurotoxin in its serotypes A, B, C1, D, E, F and G, and they all share similar structures and modes of action. Clostridial neurotoxins represent a major group of non-cytotoxic toxin molecules, and are synthesised by the host bacterium as single polypeptides that are modified post-translatibnally by a proteolytic cleavage event to form two polypeptide chains joined together by a disulphide bond. The two chains are termed the heavy chain (H-chain), which has a molecular mass of approximately 100 kDa, and the light chain (L-chain), which has a molecular mass of approximately 50 kDa. L-chains possess a protease function (zinc-dependent endopeptidase activity) and exhibit a high substrate specificity for vesicle and/or plasma membrane associated proteins involved in the exocytic process, L-chains from different clostridial species or serotypes may hydrolyse different but specific peptide bonds in one of three substrate proteins, namely synaptobrevin, syntaxin or SNAP-25. These substrates are important components of the neurosecretory machinery. Neisseria sp., most importantly from the species N. gonorrtioeae, produce functionally similar non-cytotoxic proteases. An example of such a protease is JgA protease (see WO99/58571). it has been well documented in the art that toxin molecules may be retargeted to a cell that is not the toxin's natural target cell. When so retargeted, the modified toxin is capable of binding to a desired target cell and, following subsequent translocation into the cytosol, is capable of exerting its effect on the target cell. Said re-targeting is achieved by replacing the natural Targeting Moiety (TM) of the toxin with a different TM. In this regard, the TM is selected so that it will bind to a desired target cell, and allow subsequent passage of the modified toxin into an endosome within the target cell. The modified toxin also comprises a translocation domain to enable entry of the non-cytotoxic protease into the cell cytosol. The translocation domain can be the natural translocation domain of the toxin or it can be a different translation domain obtained from a microbia! protein with translocation activity. For example, WO94/21300 describes modified clostridial neurotoxin molecules that are capable of regulating Integral Membrane Protein (IMP) density present at the cell surface of the target cell. The modified neurotoxin molecules are thus capable of controlling cell activity (eg. glucose uptake) of the target cell. WO96/33273 and WO99/17806 describe modified clostridial neurotoxin-molecules that-target peripheral sensory afferents. The modified neurotoxin molecules are thus capable of demonstrating an analgesic effect. WOOO/10598 describes the preparation of modified clostridial neurotoxin molecules that target mucus hypersecreting cells (or neuronal cells controlling said mucus hypersecreting cells), which modified neurotoxins are capable of inhibiting hypersecretion from said cells. WO01/21213 describes modified clostridial neurotoxin molecules that target a wide range of different types of non-neuronal target cells. The modified molecules are thus capable of preventing secretion from the target cells. Additional publications in the technical field of re-targeted toxin molecules include:- WOOO/62814; WOOO/04926; US5,773,586; WO93/15766; WOOO/61192; and WO99/58571. The above-mentioned TM replacement may be effected by conventional chemical conjugation techniques, which are well known to a skilled person. In this regard, reference is made to Hermanson, G.T. (1996), Bioconjugate techniques, Academic Press, and to Wong, S.S. (1991), Chemistry of protein conjugation and cross-linking, CRC Press. Chemical conjugation is, however, often imprecise. For example, following conjugation, a TM may become joined to the remainder of the conjugate at more than one attachment site. Chemical conjugation is also difficult to control. For example, a TM may become joined to the remainder of the modified toxin at an attachment site on the protease component and/ or on the translocation component. This is problematic when attachment to only one of said components (preferably at a single site) is desired for therapeutic efficacy. Thus, chemical conjugation results in a mixed population of modified toxin molecules, which is undesirable. As an alternative to chemical conjugation, TM replacement may be effected by recombinant preparation of a single polypeptide fusion protein (see WO98/07864). This technique is based on the in vivo bacterial mechanism by which native clostridial neurotoxin (ie. holotoxin) is prepared, and results in a fusion protein having the following structural arrangement: NH2 - [protease component] - [translocation component] - [TM] - COOH According to WO98/07864, the TM is placed towards the C-terminal end of the fusion protein. The fusion protein is then activated by treatment with a protease, which cleaves at a site between the protease component and the translocation component. A di-chain protein is thus produced, comprising the protease component as a single polypeptide chain covalently attached (via a disulphide bridge) to another single polypeptide chain containing the translocation component plus TM. Whilst the WO 98/07864 methodology follows (in terms of structural arrangement of the fusion protein) the natural expression system of clostridial holotoxin, the present inventors have found that this system may result in the production of certain fusion proteins that have a substantially-reduced binding ability for the intended target cell. There is therefore a need for an alternative or improved system for constructing a non-cytotoxic fusion protein. The present invention addresses one or more of the above-mentioned problems by providing a single chain, polypeptide fusion protein, comprising :- a. a non-cytotoxic protease, or a fragment thereof, which protease or protease fragment is capable of cleaving a protein of the exocytic fusion apparatus in a target cell; b. a Targeting Moiety that is capable of binding to a Binding Site on the target cell, which Binding Site is capable of undergoing endocytosis to be incorporated into an endosome within the target cell; c. a protease cleavage site at which site the fusion protein is cleavable by a protease, wherein the protease cleavage site is located between the non-cytotoxic protease or fragment thereof land the Targeting Moiety;-and a translocation domain that is capable of translocating the protease or protease fragment from within an endosome, across the endosomal membrane and into the cytosol of the target cell. The WO98/07864 system works well for the preparation of conjugates having a TM that requires a C-terminal domain for interaction with a Binding Site on a target cell. In this regard, WO98/07864 provides fusion proteins having a C-terminal domain that is "free" to interact with a Binding Site on a target cell. The present inventors have found that this structural arrangement is not -suitable for all TMs. In more detail, the present inventors have found that the WO 98/07864 fusion protein system is not optimal for TMs requiring a N-terminal domain for interaction with a binding site on a target cell. This problem is particularly acute with TMs that require a specific N-terminus amino acid residue or a specific sequence of amino acid residues including the N-terminus amino acid residue for interaction with a binding site on a target cell. In contrast to WO98/07864, the present invention provides a system for preparing non-cytotoxic conjugates, wherein the TM component of the conjugate has an N-terminal domain (or an intra domain sequence) that is capable of binding to a Binding Site on a target cell. The non-cytotoxic protease component of the present invention is a non-cytotoxic protease, or a fragment thereof, which protease or protease fragment is capable of cleaving different but specific peptide bonds in one of three substrate proteins, namely synaptobrevin, syntaxin or SNAP-25, of the exocytic fusion apparatus. These substrates are important components of the neurosecretory machinery. The non-cytotoxic protease component of the present invention is preferably a neisserial IgA protease or a fragment thereof or a clostridial neurotoxin L-chain or a fragment thereof. A particularly preferred non-cytotoxic protease component is a botulinum neurotoxin (BoNT) L-chain or a fragment thereof. The translocation-component-of-the present invention-enables-translocation of the non-cytotoxic protease (or fragment thereof) into the target cell such that functional expression of protease activity occurs within the cytosol of the target cell. The translocation component is preferably capable of forming ion-permeable pores in lipid membranes under conditions of low pH. Preferably it has been found to use only those portions of the protein molecule capable of pore-formation within the endosomal membrane. The translocation component may be obtained from a microbial protein source, in particular from a bacterial or viral protein source. Hence, in one embodiment, the translocation component is a translocating domain of an enzyme, such as a bacterial toxin or viral protein. The translocation component of the present invention is preferably a clostridial neurotoxin H-chain or a fragment thereof. Most preferably it is the HN domain (or a functional component thereof), wherein HN means a portion or fragment of the H-chain of a clostridial neurotoxin approximately equivalent to the amino-terminal half of the H-chain, or the domain corresponding to that fragment in the intact H-chain. The TM component of the present invention is responsible for binding the conjugate of the present invention to a Binding Site on a target cell. Thus, the TM component is simply a ligand through which a conjugate of the present invention binds to a selected target cell. In the context of the present invention, the target cell may be any target cell, though with the proviso that the target cell is not a nociceptive sensory afferent such as a primary sensory afferent. Thus, the TM may bind to non-neuronal cells and/ or to neuronal cells. It is routine to confirm that a TM binds to a given target cell. For example, a simple radioactive displacement experiment may be employed in which tissue or cells representative of the target cell are exposed to labelled (eg. tritiated) ligand in the presence of an excess of unlabelled ligand. In such an experiment, the relative proportions of non-specific and specific binding may be assessed, thereby allowing confirmation that the ligand binds to the target cell. Optionally, the assay may include one or more binding antagonists, and the assay may further comprise observing a loss of ligand binding. Examples of this type of experiment can be found in Hulme, E.C. (1990), Receptor-binding studies, a brief outline, pp. 303-311, fn Receptor biochemistry, A Practical Approach, Ed. E.G. Hulme, Oxford University Press, The fusion proteins of the present invention generally demonstrate a reduced binding affinity (in the region of up to 100-fold) for target cells when compared with the corresponding 'free' TM. However, despite this observation, the fusion proteins of the present invention surprisingly demonstrate good efficacy. This can be attributed to two principal features. First, the non-cytotoxic protease component is catalytic - thus, the therapeutic effect of a few such molecules is rapidly amplified. Secondly, the receptors present on the target cells need only act as a gateway for entry of the therapeutic, and need not necessarily be stimulated to a level required in order to achieve a figand-receptor mediated pharmacological response. Accordingly, the fusion proteins of the present invention may be administered at a dosage that is lower that would be employed for other types of therapeutic molecules, which are typically administered at high microgram to milligram (even up to hundreds of milligram) quantities. In contrast, the fusion proteins of the present invention may be administered at much lower dosages, typically at least 10-fold lower, and more typically at 100-fold lower. The TM preferably comprises a maximum of 50 amino acid residues, more preferably a maximum of 40 amtno acid residues, particularly preferably a maximum of 30 amino acid residues, and most preferably a maximum of 20 amino acid residues, Proteinase activated receptor ligands represent a preferred group of TMs of the present invention, in particular PAR1. PARs represent a unique subtype 7-transmembrane receptor G-protein-coupled receptors in that they are proteolyticaliy modified to expose a new extracellular N-terminus, which acts as a tethered activating ligand. PAR1 agonists (such as TFLLR) have been identified that activate their cognate receptor. Parathyroid- trormone~(PTH) -also--represents-a-preferred-TM-of- -the - present invention. PTH is released by the parathyroid gland and binds to the PTH-1 receptor. This receptor has a widespread distribution but is particularly abundant in PTH target tissues, predominantly the kidney and in bone. Thus, the most preferred TMs of the present invention are:- (Table Removed) According to one embodiment of the present invention, the TM binds to a mucus-secreting cell, or to a neuronal cell controlling or directing mucus secretion. More specifically, the TM bind to (a) cells that secrete mucins, such as epithelial goblet cells and submucosal gland mucus secreting cells, (b) cells that secrete aqueous components of mucus, such as Clara cells and serous cells, or (c) cells that control or direct mucus secretion, such as "sensory-efferent" C-fibres, or NANC neural system fibres. In this regard, particular mention is made to the TMs:- VIP; betaa adrenoreceptor agonists; gastrin-releasing peptide; and calcitonin gene related peptide. Thus, according to this embodiment, said conjugates have therapeutic application in treating mucus hypersecretion, asthma, and/ or chronic obstructive pulmonary disease. In another embodiment, the TM binds to an endocrine cell. Particular mention is made here to thyroid stimulating hormone (TSH); insulin, insulin-like growth factor; TSH releasing hormone (protirelin); FSH/LH releasing hormone (gonadorelin); corticotrophin releasing hormone (CRH); and ACTH. Thus, according to this embodiment, said conjugates have therapeutic application in -treating:--endocrine-neoplasia-including-MEN-;"thyrotoxicosis-and other diseases dependent on hypersecretions from the thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other diseases dependent on anterior pituitary hypersecretion; hyperandrogenism, chronic anovulation and other diseases associated with polycystic ovarian syndrome. In another embodiment the TM binds to an inflammatory cell. Particular mention here is made to ligands (i) for mast cells, such as the C4 domain of the Fc IgE; (ii) for eosinophils, such as ligands to the C3a/C4a-R complement receptor, antigens reactive towards CR4 complement receptor; (iii) for macrophages and monocytes, such as macrophage stimulating factor, (iv) for neutrophils, such as an antigen associated with the iC3b -complement -receptor, or IL8. Thus, according to this embodiment, said conjugates have therapeutic application for treating allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eoslnophtlia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, glomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders. in another embodiment, the TM binds to an exocrine cell. Particular mention here is made to pituitary adenyl cyclase activating peptide (PACAP-38). Thus, according to this embodiment, said conjugates have therapeutic application for treating mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon. In a further embodiment, the TM binds to an immunological cell. Mention here is made to the ligands:- Epstein Barr virus fragment/surface feature. Thus, according to this embodiment, said conjugates have therapeutic application for treating myasthenia gravis, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's -disease; Siogren's syndrome,—primary- billary—cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vufgaris, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious anaemia, type 1 diabetes mellitus. In a further embodiment the TM binds to a cardiovascular cell. Mention here is made to thrombin and TRAP (thrombin receptor agonist peptide), and ligands that bind to cardiovascular endothelial cells such as GP1b surface antigen-recognising antibodies. Thus, according to this embodiment, said conjugates have therapeutic application for treating cardiovascular conditions and/ or hypertension In a further embodiment, the TM binds to a bone cell. Mention here is made to ligands that bind to osteoblasts for the treatment of a disease selected from osteopetrosis and osteoporosis include calcitonin, and to ligands that bind to osteoctasts including osteoclast differentiation factors (eg. TRANCE, or RANKL or OPGL). Thus, according to this embodiment, said conjugates have therapeutic application for treating bone conditions. Linear and cyclic integrin binding sequences are a preferred group of TMs of the present invention. Many integrins recognise the triple Arg-Gly-Asp (RGD) peptide sequence (Ruoslahti, 1996). The RGD motif is found in over 100 proteins including fibronectin, tenascin, fibrinogen and vitronectin. The RGD-integrin interaction is exploited as a conserved mechanism of cell entry by many pathogens including coxsackievirus (Roivaninen et al., 1991) and adenovirus (Mathias et al., 1994). The linear and cyclic peptide sequences, PLAEIDGIEL and CPLAEIDGIELC respectively, have been shown to bind and internalise DNA in cells expressing 09^ integrin (Schneider et al., 1999). Other TMs of the present invention include those discovered by phage display techniques, in particular those which target and are internalised by human cells 2001); LEBP-1 (QPFMQCLCLIYDASC), LEBP-2 (RNVPPIFNDVYWIAF) and LEBP-3 (VFRVRPWYQSTSQS) (Wu ef a/., 2003); CDSAFVTVDWGRSMSLC (Florea et a/., 2003); SERSMNF, YGLPHKF, PSGAARA, LPHKSMP, LQHKSMP (Writer et a/., 2004); FSLSKPP, HSMQLST and STQAMFQ peptides (Rahim et a/., 2003). The protease cleavage site of the present invention allows cleavage (preferably controlled cleavage) of the fusion protein at a position between the non-cytotoxic protease component and the TM component. It is this cleavage -reaction that converts the fusion protein from. a single chain polypeptide into a disulphide-linked, di-chain polypeptide. According to a preferred embodiment of the present invention, the TM binds via a domain or amino acid sequence that is located away from the C-terminus of the TM. For example, the relevant binding domain may include an intra domain or an amino acid sequence located towards the middle (ie. of the linear peptide sequence) of the TM. Preferably, the relevant binding domain is located towards the N-terminus of the TM, more preferably at or near to the N-terminus. In one embodiment, the single chain polypeptide fusion may include more than one proteolytic cleavage site. However, where two or more such sites exist, they are different, thereby substantially preventing the occurrence of multiple cleavage events in the presence of a single protease. In another embodiment, it is preferred that the single chain polypeptide fusion has a single protease cleavage site. The protease cleavage sequence(s) may be introduced (and/ or any inherent cleavage sequence removed) at the DNA level by conventional means, such as by site-directed mutagenesis. Screening to confirm the presence of cleavage sequences may be performed manually or with the assistance of computer software (eg. the MapDraw program by DNASTAR, Inc.). Whilst any protease cleavage site may be employed, the following are preferred:- Enterokinase (DDDDK↓) Factor Xa (IEGR↓ / IDGR j) TEV(Tobacco Etch virus) (ENLYFQ↓G) Thrombin (LVPR↓GS) PreScission (LEVLFQ↓GP). Also embraced by the term protease cleavage site is an intein, which is a self-cleaving sequence. The self-splicing reaction is controllable, for example by varying the concentration of reducing agent present. In use, the protease cleavage site is cleaved and the N-terminal region (preferably the N-terminus) of the TM becomes exposed. The resulting polypeptide has a TM with an N-terminal domain or an intra domain that is substantially free from the remainder of the conjugate. This arrangement ensures that the N-terrninal component (or intra domain) of the TM may interact directly with a Binding Site on a target cell. In a preferred embodiment, the TM and the protease cleavage site are distanced apart in the fusion protein by at most 10 amino acid residues, more preferably by at most 5 amino acid residues, and most preferably by zero amino acid residues. Thus, following cleavage of the protease cleavage site, a conjugate is provided with a TM that has an N-terminal domain that is substantially free from the remainder of the conjugate. This arrangement ensures that the N-terminal component of the Targeting Moiety may interact directly with-a-Binding-Site-on-a-target-cell: One advantage associated with the above-mentioned activation step is that the TM only becomes susceptible to N-terminal degradation once proteolytic cleavage of the fusion protein has occurred. In addition, the selection of a specific protease cleavage site permits selective activation of the polypeptide fusion into a di-chain conformation. Construction of the single-chain polypeptide fusion of the present invention places the protease cleavage site between the TM and the non-cytotoxic protease component. It is preferred that, in the single-chain fusion, the TM is located between the protease cleavage site and the translocation component. This ensures that the TM is attached to the translocation domain (ie. as occurs with native ciostridial holotoxin), though in the case of the present invention the order of the two components is reversed vis-a-vis native holotoxin. A further advantage with this arrangement is that the TM is located in an exposed loop region of the fusion protein, which has minimal structural effects on the conformation of the fusion protein. In this regard, said loop is variously referred to as the linker, the activation loop, the inter-domain linker, or just the surface exposed loop (Schiavo et a/ 2000, Phys. Rev., 80, 717-766; Turton et a/., 2002, Trends Biochem. Sci., 27, 552-558). In one embodiment, in the single chain polypeptide, the non-cytotoxic protease component and the translocation component are linked together by a disutphide bond. Thus, following cleavage of the protease cleavage site, the polypeptide assumes a di-chain conformation, wherein the protease and -translocation components remain linked together by the disulphide bond. To this end, it is preferred that the protease and translocation components are distanced apart from one another in the single chain fusion protein by a maximum of 100 amino acid residues, more preferably a maximum of 80 amino acid residues, particularly preferably by a maximum of 60 amino acid residues, and most preferably by a maximum of 50 amino acid residues. In one embodiment, the non-cytotoxic protease component forms a disulphide bond with the translocation component of the fusion protein. For example, the amino acid residue of the protease component that forms the disulphide bond is located within the last 20, preferably within the last 10 C-terminal amino acid residues of the protease component. Similarly, the amino acid residue within the translocation component that forms the second part of the disulphide bond may be located within the first 20, preferably within the first 10 N-terminal amino acid residues of the translocation component. Alternatively, in the single chain polypeptide, the non-cytotoxic protease component and the TM may be linked together by a disulphide bond. In this regard, the amino acid residue of the TM that forms the disulphide bond is preferably located away from the N-terminus of the TM, more preferably towards to C-terminus of the TM. In one embodiment, the non-cytotoxic protease component forms a disulphide bond with the TM component of the fusion protein. In this regard, the amino acid residue of the protease component that forms the disulphide bond is preferably located within the last 20, more preferably within the last 10 C-terminal amino acid residues of the protease component. Similarly, the amino acid residue within the TM component that forms the second part of the disulphide bond is preferably located within the last 20, more preferably within the last 10 C-terminal amino acid residues of the TM. The above disulphide bond arrangements have the advantage that the protease and translocation components are arranged in a manner similar to that for native clostridial neurotoxin. By way of comparison, referring to the primary amino acid sequence for native clostridial neurotoxin, the respective cysteine amino acid residues are distanced apart by between 8 and 27 amino acid residues - taken from Popoff, MR & Marvaud, J-C, 1999, Structural & genomic features of clostridial neurotoxins, Chapter 9, in The Comprehensive Sourcebook of Bacterial Protein Toxins. Ed. Alouf & Freer:- (Table Removed) Information from proteolytic strains only The fusion protein may comprise one or more purification tags, which are located N-terminal to the protease component and/ or C-terminal to the translocation component. Whilst any purification tag may be employed, the following are preferred:- His-tag (eg. 6 * histidine), preferably as a C-terminal and/ or N-terminal tag MBP-tag (maltose binding protein), preferably as an N-terminal tag GST-tag (glutathione-S-transferase), preferably as an N-terminal tag His-MBP-tag, preferably as an N-terminal tag GST-MBP-tag, preferably as an N-terminal tag Thioredoxin-tag, preferably as an N-terminal tag CBD-tag (Chitin Binding Domain), preferably as an N-terminal tag. According to a further embodiment of the present invention, one or more peptide spacer molecules may be included in the fusion protein. For example, a peptide spacer may be employed between a purification tag and the rest of the fusion protein molecule (eg, between an N-terminal purification tag and a protease component of the present invention; and/ or between a C-terminal purification tag and a translocation component of the present invention). A peptide spacer may be also employed between the TM and translocation components of the present invention. In accordance with a second aspect of the present invention, there is provided a DMA sequence that encodes the above-mentioned single chain polypeptide. In a preferred aspect of the present invention, the DMA sequence is prepared as part of a DNA vector, wherein the vector comprises a promoter and terminator. A variety of different spacer molecules may be employed in any of the fusion proteins of the present invention. Examples of such spacer molecules include GS15, GS20, GS25, and Hx27. The present inventors have unexpectedly found that the fusion proteins of the present invention may demonstrate an improved binding activity for target cells when the size of the spacer is selected so that (in use) the C-terminus of the TM and the N-terminus of the translocation component are separated from one another by 40-105 angstroms, preferably by 50-100 angstroms, and more preferably by 50-90 angstroms. In another embodiment, the preferred spacers have an amino acid sequence of 11-29 amino acid residues, preferably 15-27 amino acid residues, and more preferably 20-27 amino acid residues. Suitable spacers may be routinely identified and obtained according to Crasto, CJ. and Feng, J.A. (2000) May; 13(5); pp. 309-312 - see also http://www.fccc./edu/re8earch/labs/fenq/limker.html. In a preferred embodiment, the vector has a promoter selected from: Promoter Induction agent Typical induction condition tac (hybrid) IPTG 0.2 mM (0.05 - 2.0 mM) AraBAD L-arabinose 0.2% (0.002 - 0.4 %) T7-/ac operator IPTG 0.2 mM (0.05 - 2.0 mM) The DNA construct of the present invention is preferably designed in silico, and then synthesised by conventional DNA synthesis techniques. The above-mentioned DNA sequence information is optionally modified for codon-biasing according to the ultimate host cell (eg. E. coli) expression system that is to be employed. The DNA backbone is preferably screened for any inherent nucleic acid sequence, which when transcribed and translated would produce an amino acid sequence corresponding to the protease cleave site encoded by the second peptide-coding sequence. This screening may be performed manually or with the assistance of computer software (eg. the MapDraw program by DNASTAR, Inc.). According to a further embodiment of the present invention, there is provided a method of preparing a non-cytotoxic agent, comprising:- a. contacting a single-chain polypeptide fusion protein of the invention with a protease capable of cleaving the protease cleavage site; b. cleaving the protease cleavage site, and thereby forming a di-chain fusion protein. This aspect provides a di-chain polypeptide, which generally mimics the structure of clostridial holotoxin. In more detail, the resulting di-chain polypeptide typically has a structure wherein:- a. the first chain comprises the non-cytotoxic protease, or a fragment thereof, which protease or protease fragment is capable of cleaving a protein of the exocytic fusion apparatus of a target cell; b. the second chain comprises the TM and the translocation domain that is capable of translocating the protease or protease fragment from within an endosome, across the endosomal membrane and into the cytosol of the target cell; and the first and second chains are disulphide linked together. According to a further aspect of the present invention, there is provided use of a single chain or di-chain polypeptide of the invention, for the manufacture of a medicament for treating, preventing or ameliorating a medical condition selected from the group consisting of mucus hypersecretion, asthma, and/ or chronic obstructive pulmonary disease, endocrine neoplasia including MEN, thyrotoxicosis and other diseases dependent on hypersecretions from the thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other diseases dependent on anterior pituitary hypersecretion; hyperandrogenism, chronic anovulation and other diseases associated with polycystic ovarian syndrome, allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eosinophilia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, giomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders, mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon, myasthenia gravis, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's disease, Sjogren's syndrome, primary biliary cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vulgaris, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious anaemla, type 1 diabetes mellitus, cardiovascular conditions and/ or hypertension, and bone conditions such as osteopetrosis and osteoporosis. According to a related aspect, there is provided a method of treating, preventing or ameliorating a medical condition or disease in a subject, comprising administering to said patient a therapeutically effective amount of a single chain or di-chain polypeptide of the invention, wherein the medical condition or disease is selected from the group consisting of mucus hypersecretion, asthma, and/ or chronic obstructive pulmonary disease, endocrine neoplasia including MEN, thyrotoxicosis and other diseases dependent on hypersecretions from the thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other diseases dependent on anterior pituitary hypersecretion; hyperandrogenism, chronic anovulation and other diseases associated with polycystic ovarian syndrome, allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eosinophilia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, glomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders, mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon, myasthenia gravis, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's disease, Sjogren's syndrome, primary biliary cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vulgarte, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious "anaemia, type 1- diabetes mellitus, cardiovascular conditions and/ or hypertension, and bone conditions such as osteopetrosis and osteoporosis. In use, the polypeptides of the present invention are typically employed in the form of a pharmaceutical composition in association with a pharmaceutical carrier, diluent and/or excipient, although the exact form of the composition may be tailored to the mode of administration. Administration is preferably to a mammal, more preferably to a human. The polypeptides may, for example, be employed in the form of an aerosol or nebulisable solution for inhalation or a sterile solution for parenteral administration, intra-articular administration or intra-cranial administration. For treating endocrine targets, i.v. injection, direct injection into gland, or aerosolisation for lung delivery are preferred; for treating inflammatory cell targets, i.v. injection, sub-cutaneous injection, or surface patch administration or aerosolisation for lung delivery are preferred; for treating exocrine targets, i.v. injection, or direct injection into or direct administration to the gland or aerosolisation for lung delivery are preferred; for treating immunological targets, i.v. injection, or injection into specific tissues eg. thymus, bone marrow, or lymph tissue are preferred; for treatment of cardiovascular targets, i.v. injection is preferred; and for treatment of bone targets, i.v. injection, or direct injection is preferred. In cases of i.v. injection, this should also include the use of pump systems. In the case of compositions for treating neuronal targets, spinal injection (eg. epidural or intrathecal) or indwelling pumps may be used. The dosage ranges for administration of the polypeptides of the present invention are those to produce the desired therapeutic effect. It will be appreciated that the dosage range required depends on the precise nature of the components, the route of administration, the nature of the formulation, the age of the patient, the nature, extent or severity of the patient's condition, contraindications, if any, and the judgement of the attending physician. Suitable daily dosages-are in-the-range-0.0004-1 mg/kg, preferably 0.0001-0.5mg/kg, more preferably 0.002-0.5mg/kg, and particularly preferably 0.004-0.5mg/kg. The unit dosage can vary from less that 1 microgram to 30mg, but typically will be in the region of 0.01 to 1mg per dose, which may be administered daily or preferably less frequently, such as weekly or six monthly. A particularly preferred dosing regimen is based on 2.5ng of fusion protein as the 1X dose per kg patient. In this regard, preferred dosages are in the range 1X-100X (ie, 2.5 - 250ng). This dosage range is significantly lower (ie. at least 10-fold, typically 100-fold lower) than would be employed with other types "of therapeutic molecules: Moreover, the above-mentioned difference is significantly magnified when the same comparison is made on a molar basis -this is because the fusion proteins of the present invention have a considerably greater molecular weight than the conventional 'small' molecule therapeutics. Wide variations in the required dosage, however, are to be expected depending on the precise nature of the components, and the differing efficiencies of various routes of administration. For example, oral administration would be expected to require higher dosages than administration by intravenous injection. Variations in these dosage levels can be adjusted using standard empirical routines for optimisation, as is well understood in the art. Compositions suitable for injection may be in the form of solutions, suspensions or emulsions, or dry powders which are dissolved or suspended in a suitable vehicle prior to use. Fluid unit dosage forms are typically prepared utilising a pyrogen-free sterile vehicle. The active ingredients, depending on the vehicle and concentration used, can be either dissolved or suspended in the vehicle. -Solutions may be-used-for-all-forms of-parenteral administration, and are particularly used for intravenous injection. In preparing solutions the components can be dissolved in the vehicle, the solution being made isotonlc if necessary by addition of sodium chloride and sterilised by filtration through a sterile filter using aseptic techniques before filling into suitable sterile vials or ampoules and sealing. Alternatively, If solution stability is adequate, the solution in its sealed containers may be sterilised by autoclaving. Advantageously additives such as buffering, solubilising, stabilising, preservative or bactericidal, suspending or emulsifying agents and/or local anaesthetic agents may be dissolved in the vehicle. Dry powders which are dissolved or suspended in a suitable vehicle prior to use may be prepared by filling pre-sterilised drug substance and other ingredients into a sterile container using aseptic technique in a sterile area. Alternatively the components (ie. agent plus inhibitor) and other ingredients may be dissolved in an aqueous vehicle, the solution is sterilized by filtration and distributed into suitable containers using aseptic technique in a sterile area. The product is then freeze dried and the containers are sealed asepticafly. Parenteral suspensions, suitable for intramuscular, subcutaneous or intradermal injection, are prepared in substantially the same manner, except that the sterile components are suspended in the sterile vehicle, instead of being dissolved and sterilisation cannot be accomplished by filtration. The components may be isolated in a sterile state or alternatively it may be sterilised after isolation, eg. by gamma irradiation. Advantageously, a suspending agent for example polyvinylpyrrolidone is included in the composition/s to facilitate uniform distribution of the components. Compositions suitable for administration via the respiratory tract include aerosolsrnebulisable-soltitions-or--microfine- powders-for insufflation. In the latter case, particle size of less than 50 microns, especially less than 10 microns, is preferred. Such compositions may be made up in a conventional manner and employed in conjunction with conventional administration devices. Definitions Section Targeting Moiety (TM) means any chemical structure associated with an agent that functionally interacts with a Binding Site to cause a physical association between the agent and the surface of a target cell. In the context of the present invention, the target cell is any cell except a nociceptive sensory afferent. The term TM embraces any molecule (ie. a naturally occurring molecule, or a chemically/physically modified variant thereof) that is capable of binding to a Binding Site on the target cell, which Binding Site is capable of internalisation (eg. endosome formation) - also referred to as receptor-mediated endocytosis. The TM may possess an endosomal membrane translocation function, in which case separate TM and Translocation Domain components need not be present in an agent of the present invention. The TM of the present invention binds (preferably specifically binds) to a target cell. The term non-cytotoxic means that the protease molecule in question does not kill the target cell to which it has been re-targeted. The protease of the present invention embraces ail naturally-occurring non-cytotoxic proteases that are capable of cleaving one or more proteins of the exocytic fusion apparatus in eukaryotic cells. The protease of the present invention is preferably a bacterial protease (or -fragment thereof). -More preferably-the-bacteral-protease-is-selected from the genera Clostridium or Neisseria (eg. a clostridia! L-chain, or a neisserial IgA protease preferably from N. gononhoeao). The present Invention also embraces modified non-cytotoxic proteases, which include amino acid sequences that do not occur in nature and/or synthetic amino acid residues, so long as the modified proteases still demonstrate the above-mentioned protease activity. The protease of the present invention preferably demonstrates a serine or metalloprotease activity (eg. endopeptidase activity). The protease is preferably specific for a SNARE protein (eg. SNAP-25, synaptobrevinA/AMP, or syntaxin). Particular mention is made to the protease domains of neurotoxins, for example the protease domains of bacterial neurotoxins. Thus, the present invention embraces the use of neurotoxin domains, which occur in nature, as well as recombinantly prepared versions of said naturally-occurring neurotoxins. Exemplary neurotoxins are produced by clostridia, and the term clostridial neurotoxin embraces neurotoxins produced by C. tetani (TeNT), and by C. botulinum (BoNT) serotypes A-G, as well as the closely related BoNT-like neurotoxins produced by C. baratii and C. butyricum. The above-mentioned abbreviations are used throughout the present specification. For example, the nomenclature BoNT/A denotes the source of neurotoxin as BoNT (serotype A). Corresponding nomenclature applies to other BoNT serotypes. The term L-chain fragment means a component of the L-chain of a neurotdxin, which fragment demonstrates a metalloprotease activity and is capable of proteolytically cleaving a vesicle and/ or plasma membrane associated protein involved In cellular exocytosis. -A-Translocation-Domain is a molecule that enables-translocation of a protease (or fragment thereof) into a target cell such that a functional expression of protease activity occurs within the cytosol of the target cell. Whether any molecule (eg. a protein or peptide) possesses the requisite translocation function of the present invention may be confirmed by any one of a number of conventional assays. For example, Shone C. (1987) describes an in vitro assay employing liposomes, which are challenged with a test molecule. Presence of the requisite translocation function is confirmed by release from the liposomes of K+ and/or labelled NAD, which may be readily monitored [see Shone C. (1987) Eur. J. Biochem; vol: 167(1): pp. 175-180]. A further example is provided by Blaustein R. (1987), which describes a simple in vitro assay employing planar phospholipid bilayer membranes. The membranes are challenged with a test molecule and the requisite translocation function is confirmed by an increase in conductance across said membranes [see Blaustein (1987) FEBS Letts; vol. 226, no. 1: pp. 115-120]. Additional methodology to enable assessment of membrane fusion and thus identification of Translocation Domains suitable for use in the present invention are provided by Methods in Enzymology Vol 220 and 221, Membrane Fusion Techniques, Parts A and B, Academic Press 1993. The Translocation Domain is preferably capable of formation of ion-permeable pores in lipid membranes under conditions of low pH. Preferably it has been found to use only those portions of the protein molecule capable of pore-formation within the endosomal membrane. The Translocation Domain may be obtained from a microbial protein source, in particular from a bacterial or viral protein source. Hence, in one embodiment, the Translocation Domain is a translocating domain of an enzyme, such as a bacterial toxin or viral protein. lt is well-documented-that-certain-domains-of-bacterial toxin-molecules are capable of forming such pores. It is also known that certain translocation domains of virally expressed membrane fusion proteins are capable of forming such pores. Such domains may be employed in the present invention. The Translocation Domain may be of a clostridial origin, namely the HN domain (or a functional component thereof). HN means a portion or fragment of the H-chain of a clostridial neurotoxin approximately equivalent to the amino-terminal half of the H-chain, or the domain corresponding to that fragment in the intact H-chain. It is preferred that the H-chain substantially lacKs the natural binding function of the Hc component of the H-chain. In this regard, the He function may be removed by deletion of the Hc amino acid sequence (either at the DNA synthesis level, or at the post-synthesis level by nuclease or protease treatment). Alternatively, the Hc function may be inactivated by chemical or biological treatment. Thus, the H-chain is preferably incapable of binding to the Binding Site on a target cell to which native clostridial neurotoxin (ie. holotoxin) binds. In one embodiment, the translocation domain is a HN domain (or a fragment thereof) of a clostridial neurotoxin. Examples of suitable clostridial Translocation Domains include:- Botulinum type A neurotoxin - amino acid residues (449-871) Botulinum type B neurotoxin - amino acid residues (441-858) Botulinum type C neurotoxin - amino acid residues (442-866) Botulinum type D neurotoxin - amino acid residues (446-862) Botulinum type E neurotoxin - amino acid residues (423-845) Botulinum type F neurotoxin - amino acid residues (440-864) Botulinum type G neurotoxin - amino acid residues (442-863) Tetanus neurotoxin - amino acid residues (458-879) For further details on the genetic basis of toxin production in Clostridium botulinum and C. tetani, we refer to Henderson et al (1997) in The Clostridia: Molecular Biology and-Pathogenesis, Academic-press:- The term HN embraces naturally-occurring neurotoxin HN portions, and modified HN portions having amino acid sequences that do not occur in nature and/or synthetic amino acid residues, so long as the modified HN portions still demonstrate the above-mentioned translocation function. Alternatively, the Translocation Domain may be of a non-clostridial origin (see Table 1), Examples of non-clostridial Translocation Domain origins include, but not be restricted to, the translocation domain of diphtheria toxin [O=Keefe et al., Proc. Natl. Acad. Sci, USA (1992) 89, 6202-6206; Silverman et a/., J. Biol. Chem. (1993) 269, 22524-22532; and London, E. (1992) Biochem. Biophys. Acta., 1112, pp.25-51], the translocation domain of Pseudomonas exotoxin type A [Prior et al. Biochemistry (1992) 31, 3555-3559], the translocation domains of anthrax toxin [Blanke et al. Proc. Natl. Acad. Sci. USA (1996) 93, 8437-8442], a variety of fusogenic or hydrophobic peptides of translocating function [Plank et al. J. Biol. Chem. (1994) 269, 12918-12924; and Wagner et al (1992) PNAS, 89, pp.7934-7938], and amphiphilic peptides [Murata et al (1992) Biochem., 31, pp. 1986-1992]. The Translocation Domain may mirror the Transiocation Domain present in a naturally-occurring protein, or may include amino acid variations so long as the variations do not destroy the translocating ability of the Translocation Domain. Particular examples of viral Translocation Domains suitable for use in the present invention include certain translocating domains of virally expressed membrane fusion proteins. For example, Wagner et a/. (1992) and Murata et a/. (1992) describe the translocation (ie. membrane fusion and vesiculation) function of a number of fusogenic and amphiphilic peptides derived from the N-terminal region of influenza virus haemagglutinin. Other virally expressed membrane fusion proteins known to have the desired translocating activity are a translocating domain of a fusogenic peptide of Semliki Forest Virus (SFV), a translocating domain of vesicular stomatitis virus (VSV) glycoprotein G, a translocating domain of SER virus F protein and a translocating domain of Foamy virus envelope glycoprotein. Virally encoded Aspike proteins have particular-application-in-the-context-of-the-present invention for-example, the E1 protein of SFV and the G protein of the G protein of VSV. Use of the Translocation Domains listed in Table 1 includes use of sequence variants thereof. A variant may comprise one or more conservative nucleic acid substitutions and/ or nucleic acid deletions or insertions, with the proviso that the variant possesses the requisite translocating function. A variant may also comprise one or more amino acid substitutions and/ or amino acid deletions or insertions, so long as the variant possesses the requisite translocating function. Table 1 (Table Removed) SEQ ID NOs SEQ ID1 DMA sequence of the LC/A SEQ ID2 DNA sequence of the HN/A SEQ ID3 DNA sequence of the LC/B SEQ ID4 DNA sequence of the HN/B SEQ-ID5 -DNA sequenee-of-the LC/C- SEQ ID6 DNA sequence of the HN/C SEQ ID7 DNA sequence of the CP PAR1-B linker SEQ ID8 DNA sequence of the CP PTH-C linker SEQ ID9 DNA sequence of the CP PAR1-B fusion SEQ ID10 Protein sequence of the CP PAR1-B fusion SEQ ID11 DNA sequence of the CP PTH-C fusion SEQ 1012 Protein sequence of the CP PTH-C fusion SEQ ID13 DNA sequence of the CP RGD-C linker SEQ 1D14 DNA sequence of the CP RGD-C fusion SEQ ID15 Protein sequence of the CP RGD-C fusion SEQ ID16 DNA sequence of the CP cyclicRGD-C linker SEQ ID17 DNA sequence of the CP cyclicRGD-C fusion SEQ ID18 Protein sequence of the CP cyclicRGD-C fusion SEQ ID19 DNA sequence of the CP THALWHT-C linker SEQ 1D20 DNA sequence of the CP THALWHT-C fusion SEQ ID21 Protein sequence of the CP THALWHT-C fusion SEQ ID22 DNA sequence of the CP cyclicTHALWHT-C linker SEQ ID23 DNA sequence of the CP cyclicTHALWHT-C fusion SEQ ID24 Protein sequence of the CP cyclicTHALWHT-C fusion SEQ ID25 DNA sequence of the CP ANP-C linker SEQ ID26 DNA sequence of the CP ANP-C fusion SEQ ID27 Protein sequence of the CP ANP-C fusion SEQ ID28 DNA sequence of the CP VIP-C linker SEQ ID29 DNA sequence of the CP VIP-C fusion SEQ ID30 Protein sequence of the CP VIP-C fusion SEQ ID31 DNA sequence of the CP Gastrin releasing peptide -C linker SEQ ID32 DNA sequence of the CP Gastrin releasing peptide-C fusion SEQ ID33 Protein sequence of the CP Gastrin releasing peptide -C fusion EXAMPLES Example 1 - Preparation of LC/B and HN/B backbone clones The following procedure creates the LC and HN fragments for use as the component backbone for multidomain fusion expression. This example is based on preparation of a serotype B based clone (SEQ ID3 and SEQ ID4), though the procedures and methods are equally applicable to the other serotypes (illustrated by the sequence listing for serotype A (SEQ ID1 and SEQ ID2) and serotype C (SEQ ID5 and SEQ ID6)). Preparation of cloning and expression vectors pCR 4 (Invitrogen) is the chosen standard cloning vector chosen due to the lack of restriction sequences within the vector and adjacent sequencing primer sites for easy construct confirmation. The expression vector is based on the pMAL (NEB) expression vector, which has the desired restriction sequences within the multiple cloning site in the correct orientation for construct insertion (BamHI-Sall-Pstl-Hindlll). A fragment of the expression vector has been removed to create a non-mobilisable plasmid and a variety of different fusion tags have been inserted to increase purification options. Preparation of protease (eg. LC/B) insert The LC/B (SEQ ID3) is created by one of two ways: The DNA sequence is designed by back translation of the LC/B amino acid sequence (obtained from freely available database sources such as GenBank (accession number P10844) or Swissprot (accession locus BXB_CLOBO) using one of a variety of reverse translation software tools (for example EditSeq best E coli reverse translation (DNASTAR Inc.), or Backtranslation tool v2.0 (Entelechon)). BamHI/Sa/l recognition sequences are incorporated at the 5' and 3' ends respectively of the sequence maintaining the correct reading frame. The DNA sequence is screened (using software such as MapDraw, DNASTAR Inc.) for restriction enzyme cleavage sequences incorporated during the back translation. Any cleavage sequences that are found to be common to those required by the cloning system are removed manually from the proposed coding sequence ensuring common E. coli codon usage is maintained. E. coli codon usage is assessed by reference to software programs such as Graphical Codon Usage Analyser (Geneart), and the %GC content and codon usage ratio assessed by reference to published codon usage tables (for example GenBank Release 143, September 13 2004). This optimised DNA sequence containing the LC/B open reading frame (ORF) is then commercially synthesized (for example by Entelechon, Geneart or Sigma-Genosys) and is provided in the pCR 4 vector. The alternative method is to use PCR amplification from an existing DNA sequence with BamHI and Sall restriction enzyme sequences incorporated into the 5' and 3' PCR primers respectively. Complementary oligonucleotide primers are chemically synthesised by a Supplier (for example MWG or Sigma-Genosys) so that each pair has the ability to hybridize to the opposite strands (3' ends pointing "towards" each other) flanking the stretch of Clostridium target DNA, one oligonucleotide for each of the two DNA strands. To generate a PCR product the pair of short oligonucleotide primers specific for the Clostridium DNA sequence are mixed with the Clostridium DNA template and other reaction components and placed in a machine (the 'PCR machine') that can change the incubation temperature of the reaction tube automatically, cycling between approximately 94 °C (for denaturation), 55 °C (for oligonucleotide annealing), and 72 °C (for synthesis). Other reagents required for amplification of a PCR product include a DNA polymerase (such as Tag or Pfu polymerase), each of the four nucleotide dNTP building blocks of DNA in equimolar amounts (50-200 µM) and a buffer appropriate for the enzyme optimised for Mg2+ concentration (0.5-5 mM). The amplification product is cloned into pCR 4 using either, TOPO TA cloning for Tag PCR products or Zero Blunt TOPO cloning for Pfu PCR products (both kits commercially available from Invitrogen). The resultant clone is checked by sequencing. Any additional restriction sequences that are not compatible with the cloning system are then removed using site directed mutagenesis (for example using Quickchange (Stratagene Inc.)). Preparation of translocation (eg. HN) insert The HN/B (SEQ ID4) is created by one of two ways: The DNA sequence is designed by back translation of the HN/B amino acid sequence (obtained from freely available database sources such as GenBank (accession number P10844) or Swissprot (accession locus BXB_CLOBO)) using one of a variety of reverse translation software tools (for example EditSeq best E. coli reverse translation (DNASTAR Inc.), or Backtranslation tool v2.0 (Entelechon)). A Pstl restriction sequence added to the N-terminus and Xbal-stop codon-Hindlll to the C-terminus ensuring the correct reading frame in maintained. The DNA sequence is screened (using software such as MapDraw, DNASTAR Inc.) for restriction enzyme cleavage sequences incorporated during the back translation. Any sequences that are found to be common to those required by the cloning system are removed manually from the proposed coding sequence ensuring common E. coli codon usage is maintained. E. coli codon usage is assessed by reference to software programs such as Graphical Codon Usage Analyser (Geneart), and the %GC content and codon usage ratio assessed by reference to published codon usage tables (for example GenBank Release 143, September 13 2004). This optimised DNA sequence is then commercially synthesized (for example by Entelechon, Geneart or Sigma-Genosys) and is provided in the pCR 4 vector. The alternative method is to use PCR amplification from an existing DNA sequence with Pstl and Xbal-stop codon-H/ndlll restriction enzyme sequences incorporated into the 5' and 3' PCR primers respectively. The PCR amplification is performed as described above. The PCR product is inserted into pCR 4 vector and checked by sequencing. Any additional restriction sequences that are not compatible with the cloning system are then removed using site directed mutagenesis (for example using Quickchange (Stratagene Inc.)). Example 2 - Preparation of a LC/B-PAR1-HN/B fusion protein Preparation of linker-PAR1-spacer insert The L.C-HN linker can be designed from first principle, using the existing sequence information for the linker as the template. For example, the serotype B linker defined as the inter-domain polypeptide region that exists between the cysteines of the disulphide bridge between LC and HN within which proteolytic activation occurs. This sequence Information is freely available from available database sources such as GenBank (accession number P10844) or Swissprot (accession locus BXB_CLOBO). It is into this linker that an Enterokinase site, PAR1 and spacer are incorporated and using one of a variety of reverse translation software tools (for example EditSeq best E. co/i reverse translation (DNASTAR Inc.), or Backtranslation tool v2.0 (Entelecnon)), the DNA sequence encoding the linker-ligand-spacer region is determined. Restriction site are then incorporated into the DNA sequence and can be arranged as BamHI-Sa/l-linker-protease site-PAR1-Mhel-spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID7). It is important to ensure the correct reading frame is maintained for the spacer, PAR1 and restriction sequences and that the Xbal sequence is not preceded by the bases, TC which would result on DAM methylation. The DNA sequence is screened for restriction sequence incorporated and any additional sequences are removed manually from the remaining sequence ensuring common E. coli codon usage is maintained. E. coli codon usage is assessed by reference to software programs such as Graphical Codon Usage Analyser (Geneart), and the %GC content and codon usage ratio assessed by reference to published codon usage tables (for example GenBank Release 143, September 13 2004). This optimised DNA sequence is then commercially synthesized (for example by Entelechon, Geneart or Sigma-Genosys) and is provided in the pCR 4 vector. Preparation of the LC/B-PAR1-HN/B fusion In order to create the LC-linker-PAR1-spacer-HN construct (SEQ ID9), the pCR 4 vector encoding the linker (SEQ ID7) is cleaved with BamHI + Sa/l restriction enzymes. This cleaved vector then serves as the recipient vector for insertion and ligation of the LC/B DNA (SEQ IDS) cleaved with BamHI + Sa/l. The resulting plasmid DNA is then cleaved with Pstl + Xbal restriction enzymes and serves as the recipient vector for the insertion and ligation of the HN/B DNA (SEQ ID4) cleaved with Pstl + Xbal. The final construct contains the LC-linker-PAR1-spacer-HN ORF (SEQ ID9) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQID10. Example 3 - Preparation LC/C-PTH-HN/C fusion protein The LC-HN linker can be designed using the methods described in example two but using the C serotype linker arranged as BamHI-Sa/I-linker-protease site-PTH-Nhel-spacer-Spel-Pstl-Xbal-stop codon-H//7dlII (SEQ ID8). The LC/C-PTH-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ 106) made using the methods described in example one and constructed using methods described in example two. The final construct contains the LC-linker-PTH-spacer-HN ORF (SEQ ID 1i) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID 12. Example 4 - Preparation and purification of LC/C-RGD-HN/C fusion protein The LC-HN linker is designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sa/l-linker-protease site-RGD-Nhel-spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID 13). The LC/C-RGD-HN/C fusion is then assembled using the LC/C (SEQ ID 5) and HN/C (SEQ ID 6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC-linker- RGD-spacer-HN ORF (SEQ ID 14) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID15, The resultant expression plasmid, pMAL LC/C-RGD-HN/C is transformed into E coli BL21 for recombinant protein expression. Expression of LC/C-RGP-HN/C fusion protein Expression of the LC/C-RGD-HN/C fusion protein is achieved using the following protocol. Inoculate 100ml of modified TB containing 0.2% glucose and 100µg/ml ampicillin in a 250ml flask with a single colony from the LC/C-RGD-HN/C expression strain. Grow the culture at 37°C, 225 rpm for 16 hours. Inoculate 1L of modified TB containing 0.2% glucose and 100µg/ml ampicillin !n a 2L flask with 10ml of overnight culture. Grow cultures at 37°C until an approximate OD600nm of 0.5 is reached at which point reduce the temperature to 16°C. After 1 hour induce the cultures with 1mM IPTG and grow at 16°C for a further-16-hours. -Figure 1 demonstrates the expressed protein in E. coli as analysed by SDS-PAGE, Purification of LC/C-RGD-HN/C fusion protein Defrost falcon tube containing 25ml 50mM HEPES pH 7.2 200mM NaCI and approximately 10 g of E. coli BL21 cell paste. Sonicate the cell paste on ice 30 seconds on, 30 seconds off for 10 cycles at a power of 22 microns ensuring the sample remains cool. Spin the lysed cells at 18 000 rpm, 4°C for 30 minutes. Load the supernatant onto a 0.1 M NiSO4 charged Chelating column (20-30 ml column is sufficient) equilibrated with 50mM HEPES pH 7.2 200mM NaCI. Using a step gradient of 10 and 40mM imidazole, wash away the non-specific bound protein and elute the fusion protein with 100mM imidazole. Dialyse the eluted fusion protein against 5L of 50mM HEPES pH 7,2 200mM NaCI at 4°C overnight and measure the OD of the dialysed fusion protein. Add 1 unit of factor Xa per 100 µg fusion protein and incubate at 25°C static overnight. Load onto a 0.1 M NiSO4 charged Chelating column (20-30 ml column is sufficient) equilibrated with 50 mM HEPES pH 7.2 200mM Wash column to baseline with 50mM HEPES pH 7.2 200mM NaCl. Using a step gradient of 10 and 40mM imidazole, wash away the non-specific bound protein and elute the fusion protein with 100mM imidazole. Dialyse the eluted fusion protein against 5L of 50mM HEPES pH 7.2 200mM NaCl at 4°C overnight and concentrate the fusion to about 2 mg/ml, aliquot sample and freeze at -20°C. Test purified protein using OD, BCA and purity analysis. Figure 2 demonstrates the purified protein as analysed by SDS-PAGE. Example 6 - Preparation LC/C-cyclicRGD-HN/C fusion protein The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sa/l-linker-protease site-cylicRGD-Nhel~spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID16) The LC/C-cyclicRGD-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC-linker-cyclicRGD-spacer-HN ORF (SEQ 1017) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID18. The resultant expression plasmid, pMAL LC/C-cyclicRGD-HN/C was transformed into E. coli BL21 for recombinant protein expression. Expression of the fusion protein was carried out as described in Example 4. Figure 1 demonstrates the protein expressed in E. coli as analysed by SDS-PAGE. Example 6 - Preparation LC/C-THALWHT-HN/C fusion protein The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype (inker arranged as BamHI-Sall-linker-protease site-THALWHT-Nhel-spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID19). The LC/C-THALWHT-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC~linker-THALWHT-spacer-HN ORF (SEQ ID20) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID21. Expression of the fusion protein was carried out as described in Example 4. Figure 1 demonstrates the protein expressed in E. coli as analysed by SDS-PAGE. The-THALWHT peptide sequence given in this Example (SEQ IDs 19, 20 and 21) can be exchanged with another peptide sequence found by phage display techniques. For example, LEBP-1 (QPFMQCLCLIYDASC), LEBP-2 (RNVPP1FNDVYWIAF) and LEBP-3 (VFRVRPWYQSTSQS) (Wu ef a/., 2003); CDSAFVTVDWGRSMSLC (Florea ef a/., 2003); SERSMNF, YGLPHKF, PSGAARA, LPHKSMP, LQHKSMP (Writer ef a/., 2004); FSLSKPP, HSMQLST and STQAMFQ peptides (Rahim ef a/., 2003). Example 7 - Preparation LC/C-cyclicTHALWHT-HN/C fusion protein The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sall-linker-protease site-cyclicTHALWHT-Nhel-spacer-Spel-Pstl-X6al-stop codon-Hindlll (SEQ ID22). The LC/C-cyciicTHALWHT-Hn/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in example one and constructed using methods described in Example 2. The final construct contains the LC-linker-cyclicTHALWHT-spacer-HN ORF (SEQ ID23) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ 1D24. Expression of the fusion protein was carried out as described in Example 4. Figure 1 demonstrates the protein expressed in E. coli as analysed by SDS-PAGE. The THALWHT peptide sequence given in this Example (SEQ IDs 19, 20 and 21) can be exchanged with another peptide sequence found by phage display techniques. For example, LEBP-1 (QPFMQCLCLIYDASC), LEBP-2 (RNVPPIFNDVYWIAF) and LEBP-3 (VFRVRPWYQSTSQS) (Wu ef a/., 2003); CDSAFVTVDWGRSMSLC (Florea ef a/., 2003); SERSMNF, YGLPHKF, PSGAARA, LPHKSMP, LQHKSMP (Writer ef a/., 2004); FSLSKPP, HSMQLST and STQAMFQ peptides (Rahim ef a/., 2003). Example 8 ~ Preparation LC/C-ANP-HN/C fusion protein The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sa/l-linker-protease site-ANP-Nhel-spacer-Spel-Pstl-Xbal-stop eodon-Hindlll (SEQ ID25). The LC/C-ANP-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC-linker-ANP-spacer-HN ORF (SEQ ID26) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID27. Example 9 - Preparation LC/C-VIP-HN/C fusion protein The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sall-linker-protease site-VIP-Nhel-spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID28). The LC/C-VIP-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC-linker-VIP-spacer-HN ORF (SEQ ID29) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID30. The VIP sequence given in SEQ ID28 could be replaced with VIP analogue or agonist sequences. For example, [R15'20'21, L17]-VIP or [R15-20-21, L17]-VIP-GRR (Kashimoto et al,, 1996; Onoue et a/., 2004), [A2,8'9'16l19'24]-VIP or Example 10 - Preparation LC/C-gastrin releasing peptide-HN/C fusion protein The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sa/l-linker-protease site-gastrin releasing peptide-Nhel-spacer-Spel-Psfl-X6al-stop codon-Hindlll (SEQ ID34). The LC/C-gastrin releasing peptide-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods-described-in Example 1-and-construeted-using methods described in Example 2. The final construct contains the LC-linker-gastrin releasing peptide-spacer-HN ORF (SEQ ID35) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID36. Example 11 - Assessment of functionality of the LC/C-RGD-HN/C fusion protein The functionality of the TM component of the LC/C-RGD-HN/C fusion protein (prepared according to Example 4) is assessed by a ligand binding assay. To facilitate assessment of ligand binding, an RGD binding-peptide is synthesised in a biotinylated and non-biotinylated form. Binding of the fusion protein is determined by a competition assay with the biotinylated form. Briefly, NCI-H292 cells are plated into 96 well plates and viable cultures established. Cells and solutions are are-chilled to 4°C and solutions are prepared in cell feeding medium-plus-HEPES (50mM). Prior to treatment, media is removed from the cells and replaced with media-plus-HEPES (500µI per well), which is then also removed. Labelled ligand, at x 2 the required concentration, is added to all wells (50µl per well). The fusion protein, at x 2 the required concentration, is then added to wells (50µl per well). After 1 hour at 4°C, the media is removed and replaced with media + HEPES (100µl per well). This media is removed and replaced with media + HEPES (100µl per well). Cells are lysed with 100µl per well PBS-Tween 0.1% for 5 mins at 4°C. PBS-Tween is removed and cells are washed with media + HEPES (100µl per well). This media is removed and replaced with 100µl PBS + 100µl streptavidin-HRP per well. Cells are incubated at RTP for 20 mins. The PBS + streptavidin is removed and the cells are washed with PBS-Tween. 100µl per well of TMB is added and the cells are incubated at 37°C for 10mins. 50µl per well 2M H2SO4 is and the plate read at 450nm. Using this methodology, the ability of the TM component of the LC/C-RGD-HN/C fusion protein to bind to the cell surface is confirmed. Description of the Figures: Figure 1 • Expression of LC/C-RGD-HN/C, LC/C-cyclicRGD-HN/C, LC/C-THALWHT-HN/C and LC/C-cycllcTHALWHT-HN/C fusion proteins in E. coll. Using the methodology outlined in Example 4, LC/C-RGD-HN/C, LC/C-cyclic RGD-HN/C, LC/C-THALWHT-HN/C and LC/C-cyclicTHALWHT-HN/C fusion proteins were expressed in E. coli BL21 cells. Briefly, 1L of TB media containing 0.2% glucose and 100pg/ml ampicillin was inoculated with 10ml of starter culture. Cultures were grown at 37°C until an approximate OD600nm of 0.5 was reached at which point the temperature was reduced to 16°C. After 1 hour the cultures were induced with 1mM IPTG and grown for a further 16 hours. Lane 1, LC/C-THALWHT-HN/C; Lane 3, LC/C-cyclicTHALWHT-HN/C; Lane 4t LC/C-cyclicRGD-HN/C. Figure 2 - Purification of a LC/C-RGD-HN/C fusion protein Using the methodology outlined in Example 5, a LC/C-RGD-HN/C fusion protein was purified from E. coli BL21 cells. Briefly, the soluble products obtained following cell disruption were applied to a nickel-charged affinity capture column. Bound proteins were eluted with 100mM imidazole, treated with Factor Xa to activate the fusion protein and remove the maltose-binding protein (MBP) tag, then re-applied to a second nickel-charged affinity capture column. Samples from the purification procedure were assessed by SDS-PAGE. The final purified material in the absence and presence of reducing agent is identified in the lanes marked [-] and [+] respectively. References Florea etal., (2003) J. Drug Targeting 11: 383-390 Jost et al., (2001) FEBS lett. 489: 263-269 Lee et al., (2001) Eur. J. Biochem. 268: 2004-2012 Mathias et al., (1994) J. Virol. 68: 6811-6814 Rahim et al., (2003) Biotechniques 35: 317-324 Roivaninen et a/., (1991) J. ViroL 65:4735-4740 Ruoslahti (1996) Ann. Rev. Ceil Dev. Biol. 12: 697-715 Schneider etal., (1999) FEBS left. 458: 329-332 Writer et a/., (2004) J. Drug Targeting 12:185-193 Wu et a/., (2003) Gene Ther. 10:1429-1436 We Claims 1. A single chain, polypeptide fusion protein, comprising: (a) a non-cytotoxic protease, or a fragment thereof, which protease or protease fragment is capable of cleaving a protein of the exocytic fusion apparatus of a target cell; (b) a Targeting Moiety that is capable of binding to a Binding Site on the target cell, which Binding Site is capable of undergoing endocytosis to be incorporated into an endosome within the target cell; (c) a protease cleavage site at which site the fusion protein is cleavable by a protease, wherein the protease cleavage site is located between the non-cytotoxic protease or fragment thereof and the Targeting Moiety; and (d) a translocation domain that is capable of translocating the protease or protease fragment from within an endosome, across the endosomal membrane and into the cytosol of the target cell; wherein the Targeting Moiety is located between the protease cleavage site and the translocation domain. 2. The fusion protein according to Claim 1, wherein the Targeting Moiety and the protease cleavage site are separated by at most 10 amino acid residues, preferably by at most 5 amino acid residues, and more preferably by at most zero amino acid residues. 3. The fusion protein according to any preceding claim, wherein the non-cytotoxic protease is a clostridial neurotoxin L-chain. 4. The fusion protein according to any preceding claim, wherein the translocation domain is the HN domain of a clostridial neurotoxin. 5. The fusion protein according to any preceding claim, wherein the Targeting Moiety comprises at most 50 amino acid residues, preferably at most 40 amino acid residues, more preferably at least 30 amino acid residues, and most preferably at most 20 amino acid residues. 6. The fusion protein according to any preceding claim, wherein the Targeting Moiety comprises a PAR ligand, preferably the ligand PARI. 7. The fusion protein according to any of Claims 1-5, wherein the Targeting Moiety comprises a ligand that binds to PTH-1, preferably the ligand comprises a PTH peptide; or wherein the Targeting Moiety comprises a linear or cyclic integrin binding sequence, preferably a triple Arg-Gly-Asp (RGD) binding sequence. 8. The fusion protein according to any preceding claim, wherein the Targeting Moiety binds to a cell selected from the group consisting of: a mucus-secreting cell or a neuronal cell controlling or directing mucus secretion; an endocrine cell; an inflammatory cell; an exocrine cell; an immunological cell; a cardiovascular cell; or a bone cell. 9. The fusion protein according to any of Claims 1-8 wherein the Targeting Moiety comprises a ligand selected from the group consisting of: TFLLR ; PAR-1 ; PTH; VIP or a VIP analogue selected from the group consisting of [R15,20,21, L17] - VIP, [R15'20'21, L17] - VIP-GRR, [A2'8'9'16'19'24] - VIP or [A2'8'9'16'19'24'25] -VIP ; beta2 adrenoreceptor agonists; gastrin-releasing peptide (GRP); calcitonin gene related peptide (CGRP); thyroid-stimulating hormone (TSH); insulin; insulin-like growth factor; TSH-releasing hormone (protirelin); FSH/LH-releasing hormone (gonadorelin); corticotrophin releasing hormone (CRH); adenocorticotrophic hormone (ACTH); pituitary adenyl cyclase activating peptide; a ligand to the C4 domain of the Fc IgE; a ligand to the C3a/C4a-R complement receptor; an antigen reactive towards the CR4 complement receptor; macrophage stimulating factor; an antigen associated with the iC3b complement receptor; IL8; Epstein Barr virus fragment/surface feature ; thrombin; thrombin receptor agonist peptide (TRAP); GP1b surface antigen-recognising antibodies ; calcitonin ; the osteoclast differentiation factor TRANCE, RANKL, or OPGL ; a linear or cyclic peptide selected from the group consisting of: THALWHT, LEBP-1 (QPFMQCLCLIYDASC), LEBP-2 (RNVPPIFNDVYWIAF), LEBP-3 (VFRVRPWYQSTSQS), CDSAFVTVDWGRSMSLC, SERSMNF, YGLPHKF, PSGAARA, LPHKSMP, LQHKSMP, FSLSKPP, HSMQLST, STQAMFQ ; and ANP. 10. The fusion protein according to any preceding claim, wherein the fusion protein comprises a purification tag. 11. The fusion according to Claim 10, wherein the fusion protein comprises a purification tag, which is present at the N-terminal and/ or C-terminal end of the fusion protein. 12. The fusion protein according to Claims 10 or 11, wherein the purification tag is joined to the fusion protein by a peptide spacer molecule. 13. The fusion protein according to any preceding claim, wherein the translocation domain is separated from the Targeting Moiety by a peptide spacer molecule. 14. A polypeptide fusion protein comprising any one of SEQ ID NOs: 10, 12, 15, 18, 21,24, 27, 30, and 33. 15. A nucleic acid sequence encoding the polypeptide fusion protein according to any preceding Claim. 16. A nucleic acid sequence according to Claim 15, wherein the nucleic acid molecule comprises any one of SEQ ID NOs: 1-9, 11, 14, 17, 20, 23, 26, 29, and 32. 17. A DNA vector, which comprises a promoter, a nucleic acid sequence according to Claim 15 or Claim 16, wherein said DNA sequence is located downstream of the promoter, and a terminator is located downstream of the DNA construct. 18. The complementary DNA strand of the DNA sequence according to Claim 15 or Claim 16. 19. A method for preparing a single-chain polypeptide fusion protein according to any of Claims 1-14, comprising expressing a nucleic acid sequence according to Claim 15 or Claim 16, or a DNA vector according to Claim 17, in a host cell. 20. A method of preparing a non-cytotoxic agent, comprising: (a) contacting a single-chain polypeptide fusion protein according to any of Claims 1-14 with a protease capable of cleaving the protease cleavage site; (b) cleaving the protease cleavage site; and thereby forming a di-chain fusion protein. 21. A non-cytotoxic polypeptide, obtainable by the method of Claim 20, wherein the polypeptide is a di-chain polypeptide, and wherein: (a) the first chain comprises the non-cytotoxic protease, or a fragment thereof, which protease or protease fragment is capable of cleaving a protein of the exocytic fusion apparatus of a target cell; (b) the second chain comprises the TM and the translocation domain that is capable of translocating the protease or protease fragment from within an endosome, across the endosomal membrane and into the cytosol of the target cell; and the first and second chains are disulphide linked together. 22. Use of a fusion protein according to any of Claims 1-14 or a polypeptide according to Claim 21, for the manufacture of a medicament for treating, preventing or ameliorating a medical condition or disease selected from the group consisting of mucus hypersecretion, asthma, and/ or chronic obstructive pulmonary disease, endocrine neoplasia including MEN, thyrotoxicosis and other diseases dependent on hypersecretions from the thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other diseases dependent on anterior pituitary hypersecretion; hyperandrogenism, chronic anovulation and other diseases associated with polycystic ovarian syndrome, allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eosinophilia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, glomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders, mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon, myasthenia gravis, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's disease, Sjogren's syndrome, primary biliary cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vulgaris, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious anaemia, type 1 diabetes mellitus, cardiovascular conditions and/ or hypertension, and bone conditions such as osteopetrosis and osteoporosis. 23. A method of treating, preventing or ameliorating a medical condition or disease in a subject, comprising administering to said patient a therapeutically effective amount of a fusion protein according to any of Claims 1-14 or a polypeptide according to Claim 21, wherein the medical condition or disease is selected from the group consisting of mucus hypersecretion, asthma, and/ or chronic obstructive pulmonary disease, endocrine neoplasia including MEN, thyrotoxicosis and other diseases dependent on hypersecretions from the thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other diseases dependent on anterior pituitary hypersecretion; hyperandrogenism, chronic anovulation and other diseases associated with polycystic ovarian syndrome, allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eosinophilia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, glomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders, mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon, myasthenia gravis, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's disease, Sjogren's syndrome, primary biliary cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vulgaris, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious anaemia, type 1 diabetes mellitus, cardiovascular conditions and/ or hypertension, and bone conditions such as osteopetrosis and osteoporosis

Full Text

Fusion Proteins
This invention relates to non-cytotoxic fusion proteins, and to the therapeutic application thereof.
Toxins may be generally divided into two groups according to the type of effect that they have on a target cell. In more detail, the first group of toxins kill their natural target cells, and are therefore known as cytotoxic toxin molecules. This group of toxins is exemplified inter alia by plant toxins such as ricin, and abrin, and by bacterial toxins such as diphtheria toxin, and Pseudomonas exotoxin A. Cytotoxic toxins have attracted much interest in the design of "magic bullets" (eg. immunoconjugates, which comprise a cytotoxic toxin component and an antibody that binds to a specific marker on a target cell) for the treatment of cellular disorders and conditions such as cancer. Cytotoxic toxins typically kill their target cells by inhibiting the cellular process of protein synthesis.
The second group of toxins, which are known as non-cytotoxic toxins, do not (as their name confirms) kill their natural target cells. Non-cytotoxic toxins have attracted much less commercial interest than have their cytotoxic counterparts, and exert their effects on a target cell by inhibiting cellular processes other than protein synthesis. Non-cytotoxic toxins are produced by a variety of plants, and by a variety of microorganisms such as Clostridium sp. and Neisseria sp.
Clostridial neurotoxins are proteins that typically have a molecular mass of the order of 150 kDa. They are produced by various species of bacteria, especially of the genus Clostridium, most importantly C. tetani and several strains of C. botulinum, C. butyricum and C. argentinense. There are at present eight different classes of the clostridial neurotoxin, namely: tetanus toxin, and botulinum neurotoxin in its serotypes A, B, C1, D, E, F and G, and they all share similar structures and modes of action.

Clostridial neurotoxins represent a major group of non-cytotoxic toxin molecules, and are synthesised by the host bacterium as single polypeptides that are modified post-translatibnally by a proteolytic cleavage event to form two polypeptide chains joined together by a disulphide bond. The two chains are termed the heavy chain (H-chain), which has a molecular mass of approximately 100 kDa, and the light chain (L-chain), which has a molecular mass of approximately 50 kDa.
L-chains possess a protease function (zinc-dependent endopeptidase activity) and exhibit a high substrate specificity for vesicle and/or plasma membrane associated proteins involved in the exocytic process, L-chains from different clostridial species or serotypes may hydrolyse different but specific peptide bonds in one of three substrate proteins, namely synaptobrevin, syntaxin or SNAP-25. These substrates are important components of the neurosecretory machinery.
Neisseria sp., most importantly from the species N. gonorrtioeae, produce functionally similar non-cytotoxic proteases. An example of such a protease is JgA protease (see WO99/58571).
it has been well documented in the art that toxin molecules may be retargeted to a cell that is not the toxin's natural target cell. When so retargeted, the modified toxin is capable of binding to a desired target cell and, following subsequent translocation into the cytosol, is capable of exerting its effect on the target cell. Said re-targeting is achieved by replacing the natural Targeting Moiety (TM) of the toxin with a different TM. In this regard, the TM is selected so that it will bind to a desired target cell, and allow subsequent passage of the modified toxin into an endosome within the target cell. The modified toxin also comprises a translocation domain to enable entry of the non-cytotoxic protease into the cell cytosol. The translocation domain can be the natural translocation domain of the toxin or it can be a different translation domain obtained from a microbia! protein with translocation activity.

For example, WO94/21300 describes modified clostridial neurotoxin molecules that are capable of regulating Integral Membrane Protein (IMP) density present at the cell surface of the target cell. The modified neurotoxin molecules are thus capable of controlling cell activity (eg. glucose uptake) of the target cell. WO96/33273 and WO99/17806 describe modified clostridial neurotoxin-molecules that-target peripheral sensory afferents. The modified neurotoxin molecules are thus capable of demonstrating an analgesic effect. WOOO/10598 describes the preparation of modified clostridial neurotoxin molecules that target mucus hypersecreting cells (or neuronal cells controlling said mucus hypersecreting cells), which modified neurotoxins are capable of inhibiting hypersecretion from said cells. WO01/21213 describes modified clostridial neurotoxin molecules that target a wide range of different types of non-neuronal target cells. The modified molecules are thus capable of preventing secretion from the target cells. Additional publications in the technical field of re-targeted toxin molecules include:- WOOO/62814; WOOO/04926; US5,773,586; WO93/15766; WOOO/61192; and WO99/58571.
The above-mentioned TM replacement may be effected by conventional chemical conjugation techniques, which are well known to a skilled person. In this regard, reference is made to Hermanson, G.T. (1996), Bioconjugate techniques, Academic Press, and to Wong, S.S. (1991), Chemistry of protein conjugation and cross-linking, CRC Press.
Chemical conjugation is, however, often imprecise. For example, following conjugation, a TM may become joined to the remainder of the conjugate at more than one attachment site.
Chemical conjugation is also difficult to control. For example, a TM may become joined to the remainder of the modified toxin at an attachment site on the protease component and/ or on the translocation component. This is problematic when attachment to only one of said components (preferably at a single site) is desired for therapeutic efficacy.

Thus, chemical conjugation results in a mixed population of modified toxin molecules, which is undesirable.
As an alternative to chemical conjugation, TM replacement may be effected by recombinant preparation of a single polypeptide fusion protein (see WO98/07864). This technique is based on the in vivo bacterial mechanism by which native clostridial neurotoxin (ie. holotoxin) is prepared, and results in a fusion protein having the following structural arrangement:
NH2 - [protease component] - [translocation component] - [TM] - COOH
According to WO98/07864, the TM is placed towards the C-terminal end of the fusion protein. The fusion protein is then activated by treatment with a protease, which cleaves at a site between the protease component and the translocation component. A di-chain protein is thus produced, comprising the protease component as a single polypeptide chain covalently attached (via a disulphide bridge) to another single polypeptide chain containing the translocation component plus TM. Whilst the WO 98/07864 methodology follows (in terms of structural arrangement of the fusion protein) the natural expression system of clostridial holotoxin, the present inventors have found that this system may result in the production of certain fusion proteins that have a substantially-reduced binding ability for the intended target cell.
There is therefore a need for an alternative or improved system for constructing a non-cytotoxic fusion protein.
The present invention addresses one or more of the above-mentioned problems by providing a single chain, polypeptide fusion protein, comprising :-
a. a non-cytotoxic protease, or a fragment thereof, which protease
or protease fragment is capable of cleaving a protein of the
exocytic fusion apparatus in a target cell;
b. a Targeting Moiety that is capable of binding to a Binding Site on
the target cell, which Binding Site is capable of undergoing

endocytosis to be incorporated into an endosome within the
target cell; c. a protease cleavage site at which site the fusion protein is
cleavable by a protease, wherein the protease cleavage site is
located between the non-cytotoxic protease or fragment thereof
land the Targeting Moiety;-and
a translocation domain that is capable of translocating the protease or protease fragment from within an endosome, across the endosomal membrane and into the cytosol of the target cell.
The WO98/07864 system works well for the preparation of conjugates having a TM that requires a C-terminal domain for interaction with a Binding Site on a target cell. In this regard, WO98/07864 provides fusion proteins having a C-terminal domain that is "free" to interact with a Binding Site on a target cell. The present inventors have found that this structural arrangement is not -suitable for all TMs. In more detail, the present inventors have found that the WO 98/07864 fusion protein system is not optimal for TMs requiring a N-terminal domain for interaction with a binding site on a target cell. This problem is particularly acute with TMs that require a specific N-terminus amino acid residue or a specific sequence of amino acid residues including the N-terminus amino acid residue for interaction with a binding site on a target cell.
In contrast to WO98/07864, the present invention provides a system for preparing non-cytotoxic conjugates, wherein the TM component of the conjugate has an N-terminal domain (or an intra domain sequence) that is capable of binding to a Binding Site on a target cell.
The non-cytotoxic protease component of the present invention is a non-cytotoxic protease, or a fragment thereof, which protease or protease fragment is capable of cleaving different but specific peptide bonds in one of three substrate proteins, namely synaptobrevin, syntaxin or SNAP-25, of the exocytic fusion apparatus. These substrates are important components of the neurosecretory machinery. The non-cytotoxic protease component of the

present invention is preferably a neisserial IgA protease or a fragment thereof or a clostridial neurotoxin L-chain or a fragment thereof. A particularly preferred non-cytotoxic protease component is a botulinum neurotoxin (BoNT) L-chain or a fragment thereof.
The translocation-component-of-the present invention-enables-translocation of the non-cytotoxic protease (or fragment thereof) into the target cell such that functional expression of protease activity occurs within the cytosol of the target cell. The translocation component is preferably capable of forming ion-permeable pores in lipid membranes under conditions of low pH. Preferably it has been found to use only those portions of the protein molecule capable of pore-formation within the endosomal membrane. The translocation component may be obtained from a microbial protein source, in particular from a bacterial or viral protein source. Hence, in one embodiment, the translocation component is a translocating domain of an enzyme, such as a bacterial toxin or viral protein. The translocation component of the present invention is preferably a clostridial neurotoxin H-chain or a fragment thereof. Most preferably it is the HN domain (or a functional component thereof), wherein HN means a portion or fragment of the H-chain of a clostridial neurotoxin approximately equivalent to the amino-terminal half of the H-chain, or the domain corresponding to that fragment in the intact H-chain.
The TM component of the present invention is responsible for binding the conjugate of the present invention to a Binding Site on a target cell. Thus, the TM component is simply a ligand through which a conjugate of the present invention binds to a selected target cell.
In the context of the present invention, the target cell may be any target cell, though with the proviso that the target cell is not a nociceptive sensory afferent such as a primary sensory afferent. Thus, the TM may bind to non-neuronal cells and/ or to neuronal cells.
It is routine to confirm that a TM binds to a given target cell. For example, a simple radioactive displacement experiment may be employed in which tissue

or cells representative of the target cell are exposed to labelled (eg. tritiated) ligand in the presence of an excess of unlabelled ligand. In such an experiment, the relative proportions of non-specific and specific binding may be assessed, thereby allowing confirmation that the ligand binds to the target cell. Optionally, the assay may include one or more binding antagonists, and the assay may further comprise observing a loss of ligand binding. Examples of this type of experiment can be found in Hulme, E.C. (1990), Receptor-binding studies, a brief outline, pp. 303-311, fn Receptor biochemistry, A Practical Approach, Ed. E.G. Hulme, Oxford University Press,
The fusion proteins of the present invention generally demonstrate a reduced binding affinity (in the region of up to 100-fold) for target cells when compared with the corresponding 'free' TM. However, despite this observation, the fusion proteins of the present invention surprisingly demonstrate good efficacy. This can be attributed to two principal features. First, the non-cytotoxic protease component is catalytic - thus, the therapeutic effect of a few such molecules is rapidly amplified. Secondly, the receptors present on the target cells need only act as a gateway for entry of the therapeutic, and need not necessarily be stimulated to a level required in order to achieve a figand-receptor mediated pharmacological response. Accordingly, the fusion proteins of the present invention may be administered at a dosage that is lower that would be employed for other types of therapeutic molecules, which are typically administered at high microgram to milligram (even up to hundreds of milligram) quantities. In contrast, the fusion proteins of the present invention may be administered at much lower dosages, typically at least 10-fold lower, and more typically at 100-fold lower.
The TM preferably comprises a maximum of 50 amino acid residues, more preferably a maximum of 40 amtno acid residues, particularly preferably a maximum of 30 amino acid residues, and most preferably a maximum of 20 amino acid residues,
Proteinase activated receptor ligands represent a preferred group of TMs of the present invention, in particular PAR1. PARs represent a unique subtype

7-transmembrane receptor G-protein-coupled receptors in that they are proteolyticaliy modified to expose a new extracellular N-terminus, which acts as a tethered activating ligand. PAR1 agonists (such as TFLLR) have been identified that activate their cognate receptor.
Parathyroid- trormone~(PTH) -also--represents-a-preferred-TM-of- -the - present invention. PTH is released by the parathyroid gland and binds to the PTH-1 receptor. This receptor has a widespread distribution but is particularly abundant in PTH target tissues, predominantly the kidney and in bone.
Thus, the most preferred TMs of the present invention are:-

(Table Removed)
According to one embodiment of the present invention, the TM binds to a mucus-secreting cell, or to a neuronal cell controlling or directing mucus secretion. More specifically, the TM bind to (a) cells that secrete mucins, such as epithelial goblet cells and submucosal gland mucus secreting cells, (b) cells that secrete aqueous components of mucus, such as Clara cells and serous cells, or (c) cells that control or direct mucus secretion, such as "sensory-efferent" C-fibres, or NANC neural system fibres. In this regard, particular mention is made to the TMs:- VIP; betaa adrenoreceptor agonists; gastrin-releasing peptide; and calcitonin gene related peptide. Thus, according to this embodiment, said conjugates have therapeutic application in treating mucus hypersecretion, asthma, and/ or chronic obstructive pulmonary disease.

In another embodiment, the TM binds to an endocrine cell. Particular mention is made here to thyroid stimulating hormone (TSH); insulin, insulin-like growth factor; TSH releasing hormone (protirelin); FSH/LH releasing hormone (gonadorelin); corticotrophin releasing hormone (CRH); and ACTH. Thus, according to this embodiment, said conjugates have therapeutic application in
-treating:--endocrine-neoplasia-including-MEN-;"thyrotoxicosis-and other diseases dependent on hypersecretions from the thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other diseases dependent on anterior pituitary hypersecretion; hyperandrogenism, chronic anovulation and other diseases associated with polycystic ovarian syndrome.
In another embodiment the TM binds to an inflammatory cell. Particular mention here is made to ligands (i) for mast cells, such as the C4 domain of the Fc IgE; (ii) for eosinophils, such as ligands to the C3a/C4a-R complement receptor, antigens reactive towards CR4 complement receptor; (iii) for macrophages and monocytes, such as macrophage stimulating factor, (iv) for neutrophils, such as an antigen associated with the iC3b -complement -receptor, or IL8. Thus, according to this embodiment, said conjugates have therapeutic application for treating allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eoslnophtlia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, glomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders.
in another embodiment, the TM binds to an exocrine cell. Particular mention here is made to pituitary adenyl cyclase activating peptide (PACAP-38). Thus, according to this embodiment, said conjugates have therapeutic application for treating mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon.
In a further embodiment, the TM binds to an immunological cell. Mention here is made to the ligands:- Epstein Barr virus fragment/surface feature. Thus, according to this embodiment, said conjugates have therapeutic application for

treating myasthenia gravis, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's -disease; Siogren's syndrome,—primary- billary—cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vufgaris, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious anaemia, type 1 diabetes mellitus.
In a further embodiment the TM binds to a cardiovascular cell. Mention here is made to thrombin and TRAP (thrombin receptor agonist peptide), and ligands that bind to cardiovascular endothelial cells such as GP1b surface antigen-recognising antibodies. Thus, according to this embodiment, said conjugates have therapeutic application for treating cardiovascular conditions and/ or hypertension
In a further embodiment, the TM binds to a bone cell. Mention here is made to ligands that bind to osteoblasts for the treatment of a disease selected from osteopetrosis and osteoporosis include calcitonin, and to ligands that bind to osteoctasts including osteoclast differentiation factors (eg. TRANCE, or RANKL or OPGL). Thus, according to this embodiment, said conjugates have therapeutic application for treating bone conditions.
Linear and cyclic integrin binding sequences are a preferred group of TMs of the present invention. Many integrins recognise the triple Arg-Gly-Asp (RGD) peptide sequence (Ruoslahti, 1996). The RGD motif is found in over 100 proteins including fibronectin, tenascin, fibrinogen and vitronectin. The RGD-integrin interaction is exploited as a conserved mechanism of cell entry by many pathogens including coxsackievirus (Roivaninen et al., 1991) and adenovirus (Mathias et al., 1994). The linear and cyclic peptide sequences,

PLAEIDGIEL and CPLAEIDGIELC respectively, have been shown to bind and internalise DNA in cells expressing 09^ integrin (Schneider et al., 1999).
Other TMs of the present invention include those discovered by phage display techniques, in particular those which target and are internalised by human cells
2001); LEBP-1 (QPFMQCLCLIYDASC), LEBP-2 (RNVPPIFNDVYWIAF) and LEBP-3 (VFRVRPWYQSTSQS) (Wu ef a/., 2003); CDSAFVTVDWGRSMSLC (Florea et a/., 2003); SERSMNF, YGLPHKF, PSGAARA, LPHKSMP, LQHKSMP (Writer et a/., 2004); FSLSKPP, HSMQLST and STQAMFQ peptides (Rahim et a/., 2003).
The protease cleavage site of the present invention allows cleavage (preferably controlled cleavage) of the fusion protein at a position between the non-cytotoxic protease component and the TM component. It is this cleavage -reaction that converts the fusion protein from. a single chain polypeptide into a disulphide-linked, di-chain polypeptide.
According to a preferred embodiment of the present invention, the TM binds via a domain or amino acid sequence that is located away from the C-terminus of the TM. For example, the relevant binding domain may include an intra domain or an amino acid sequence located towards the middle (ie. of the linear peptide sequence) of the TM. Preferably, the relevant binding domain is located towards the N-terminus of the TM, more preferably at or near to the N-terminus.
In one embodiment, the single chain polypeptide fusion may include more than one proteolytic cleavage site. However, where two or more such sites exist, they are different, thereby substantially preventing the occurrence of multiple cleavage events in the presence of a single protease. In another embodiment, it is preferred that the single chain polypeptide fusion has a single protease cleavage site.

The protease cleavage sequence(s) may be introduced (and/ or any inherent cleavage sequence removed) at the DNA level by conventional means, such as by site-directed mutagenesis. Screening to confirm the presence of cleavage sequences may be performed manually or with the assistance of computer software (eg. the MapDraw program by DNASTAR, Inc.).
Whilst any protease cleavage site may be employed, the following are preferred:-
Enterokinase (DDDDK↓)
Factor Xa (IEGR↓ / IDGR j)
TEV(Tobacco Etch virus) (ENLYFQ↓G)
Thrombin (LVPR↓GS)
PreScission (LEVLFQ↓GP).
Also embraced by the term protease cleavage site is an intein, which is a self-cleaving sequence. The self-splicing reaction is controllable, for example by varying the concentration of reducing agent present.
In use, the protease cleavage site is cleaved and the N-terminal region (preferably the N-terminus) of the TM becomes exposed. The resulting polypeptide has a TM with an N-terminal domain or an intra domain that is substantially free from the remainder of the conjugate. This arrangement ensures that the N-terrninal component (or intra domain) of the TM may interact directly with a Binding Site on a target cell.
In a preferred embodiment, the TM and the protease cleavage site are distanced apart in the fusion protein by at most 10 amino acid residues, more

preferably by at most 5 amino acid residues, and most preferably by zero amino acid residues. Thus, following cleavage of the protease cleavage site, a conjugate is provided with a TM that has an N-terminal domain that is substantially free from the remainder of the conjugate. This arrangement ensures that the N-terminal component of the Targeting Moiety may interact directly with-a-Binding-Site-on-a-target-cell:
One advantage associated with the above-mentioned activation step is that the TM only becomes susceptible to N-terminal degradation once proteolytic cleavage of the fusion protein has occurred. In addition, the selection of a specific protease cleavage site permits selective activation of the polypeptide fusion into a di-chain conformation.
Construction of the single-chain polypeptide fusion of the present invention places the protease cleavage site between the TM and the non-cytotoxic protease component.
It is preferred that, in the single-chain fusion, the TM is located between the protease cleavage site and the translocation component. This ensures that the TM is attached to the translocation domain (ie. as occurs with native ciostridial holotoxin), though in the case of the present invention the order of the two components is reversed vis-a-vis native holotoxin. A further advantage with this arrangement is that the TM is located in an exposed loop region of the fusion protein, which has minimal structural effects on the conformation of the fusion protein. In this regard, said loop is variously referred to as the linker, the activation loop, the inter-domain linker, or just the surface exposed loop (Schiavo et a/ 2000, Phys. Rev., 80, 717-766; Turton et a/., 2002, Trends Biochem. Sci., 27, 552-558).
In one embodiment, in the single chain polypeptide, the non-cytotoxic protease component and the translocation component are linked together by a disutphide bond. Thus, following cleavage of the protease cleavage site, the polypeptide assumes a di-chain conformation, wherein the protease and

-translocation components remain linked together by the disulphide bond. To this end, it is preferred that the protease and translocation components are distanced apart from one another in the single chain fusion protein by a maximum of 100 amino acid residues, more preferably a maximum of 80 amino acid residues, particularly preferably by a maximum of 60 amino acid residues, and most preferably by a maximum of 50 amino acid residues.
In one embodiment, the non-cytotoxic protease component forms a disulphide bond with the translocation component of the fusion protein. For example, the amino acid residue of the protease component that forms the disulphide bond is located within the last 20, preferably within the last 10 C-terminal amino acid residues of the protease component. Similarly, the amino acid residue within the translocation component that forms the second part of the disulphide bond may be located within the first 20, preferably within the first 10 N-terminal amino acid residues of the translocation component.
Alternatively, in the single chain polypeptide, the non-cytotoxic protease component and the TM may be linked together by a disulphide bond. In this regard, the amino acid residue of the TM that forms the disulphide bond is preferably located away from the N-terminus of the TM, more preferably towards to C-terminus of the TM.
In one embodiment, the non-cytotoxic protease component forms a disulphide bond with the TM component of the fusion protein. In this regard, the amino acid residue of the protease component that forms the disulphide bond is preferably located within the last 20, more preferably within the last 10 C-terminal amino acid residues of the protease component. Similarly, the amino acid residue within the TM component that forms the second part of the disulphide bond is preferably located within the last 20, more preferably within the last 10 C-terminal amino acid residues of the TM.
The above disulphide bond arrangements have the advantage that the protease and translocation components are arranged in a manner similar to that for native clostridial neurotoxin. By way of comparison, referring to the

primary amino acid sequence for native clostridial neurotoxin, the respective cysteine amino acid residues are distanced apart by between 8 and 27 amino acid residues - taken from Popoff, MR & Marvaud, J-C, 1999, Structural & genomic features of clostridial neurotoxins, Chapter 9, in The Comprehensive Sourcebook of Bacterial Protein Toxins. Ed. Alouf & Freer:-

(Table Removed)
Information from proteolytic strains only
The fusion protein may comprise one or more purification tags, which are located N-terminal to the protease component and/ or C-terminal to the translocation component.
Whilst any purification tag may be employed, the following are preferred:-
His-tag (eg. 6 * histidine), preferably as a C-terminal and/ or N-terminal tag MBP-tag (maltose binding protein), preferably as an N-terminal tag GST-tag (glutathione-S-transferase), preferably as an N-terminal tag His-MBP-tag, preferably as an N-terminal tag GST-MBP-tag, preferably as an N-terminal tag Thioredoxin-tag, preferably as an N-terminal tag

CBD-tag (Chitin Binding Domain), preferably as an N-terminal tag.
According to a further embodiment of the present invention, one or more peptide spacer molecules may be included in the fusion protein. For example, a peptide spacer may be employed between a purification tag and the rest of the fusion protein molecule (eg, between an N-terminal purification tag and a protease component of the present invention; and/ or between a C-terminal purification tag and a translocation component of the present invention). A peptide spacer may be also employed between the TM and translocation components of the present invention.
In accordance with a second aspect of the present invention, there is provided a DMA sequence that encodes the above-mentioned single chain polypeptide.
In a preferred aspect of the present invention, the DMA sequence is prepared as part of a DNA vector, wherein the vector comprises a promoter and terminator.
A variety of different spacer molecules may be employed in any of the fusion proteins of the present invention. Examples of such spacer molecules include GS15, GS20, GS25, and Hx27.
The present inventors have unexpectedly found that the fusion proteins of the present invention may demonstrate an improved binding activity for target cells when the size of the spacer is selected so that (in use) the C-terminus of the TM and the N-terminus of the translocation component are separated from one another by 40-105 angstroms, preferably by 50-100 angstroms, and more preferably by 50-90 angstroms. In another embodiment, the preferred spacers have an amino acid sequence of 11-29 amino acid residues, preferably 15-27 amino acid residues, and more preferably 20-27 amino acid residues. Suitable spacers may be routinely identified and obtained according

to Crasto, CJ. and Feng, J.A. (2000) May; 13(5); pp. 309-312 - see also http://www.fccc./edu/re8earch/labs/fenq/limker.html.
In a preferred embodiment, the vector has a promoter selected from:
Promoter Induction agent Typical induction condition
tac (hybrid) IPTG 0.2 mM (0.05 - 2.0 mM)
AraBAD L-arabinose 0.2% (0.002 - 0.4 %)
T7-/ac operator IPTG 0.2 mM (0.05 - 2.0 mM)
The DNA construct of the present invention is preferably designed in silico, and then synthesised by conventional DNA synthesis techniques.
The above-mentioned DNA sequence information is optionally modified for codon-biasing according to the ultimate host cell (eg. E. coli) expression system that is to be employed.
The DNA backbone is preferably screened for any inherent nucleic acid sequence, which when transcribed and translated would produce an amino acid sequence corresponding to the protease cleave site encoded by the second peptide-coding sequence. This screening may be performed manually or with the assistance of computer software (eg. the MapDraw program by DNASTAR, Inc.).
According to a further embodiment of the present invention, there is provided a method of preparing a non-cytotoxic agent, comprising:-
a. contacting a single-chain polypeptide fusion protein of the invention with a protease capable of cleaving the protease cleavage site;

b. cleaving the protease cleavage site, and thereby forming a di-chain fusion protein.
This aspect provides a di-chain polypeptide, which generally mimics the structure of clostridial holotoxin. In more detail, the resulting di-chain polypeptide typically has a structure wherein:-
a. the first chain comprises the non-cytotoxic protease, or a
fragment thereof, which protease or protease fragment is
capable of cleaving a protein of the exocytic fusion apparatus of
a target cell;
b. the second chain comprises the TM and the translocation
domain that is capable of translocating the protease or protease
fragment from within an endosome, across the endosomal
membrane and into the cytosol of the target cell; and
the first and second chains are disulphide linked together.
According to a further aspect of the present invention, there is provided use of a single chain or di-chain polypeptide of the invention, for the manufacture of a medicament for treating, preventing or ameliorating a medical condition selected from the group consisting of mucus hypersecretion, asthma, and/ or chronic obstructive pulmonary disease, endocrine neoplasia including MEN, thyrotoxicosis and other diseases dependent on hypersecretions from the thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other diseases dependent on anterior pituitary hypersecretion; hyperandrogenism, chronic anovulation and other diseases associated with polycystic ovarian syndrome, allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eosinophilia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, giomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders, mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon, myasthenia gravis,

rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's disease, Sjogren's syndrome, primary biliary cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vulgaris, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious anaemla, type 1 diabetes mellitus, cardiovascular conditions and/ or hypertension, and bone conditions such as osteopetrosis and osteoporosis.
According to a related aspect, there is provided a method of treating, preventing or ameliorating a medical condition or disease in a subject, comprising administering to said patient a therapeutically effective amount of a single chain or di-chain polypeptide of the invention, wherein the medical condition or disease is selected from the group consisting of mucus hypersecretion, asthma, and/ or chronic obstructive pulmonary disease, endocrine neoplasia including MEN, thyrotoxicosis and other diseases dependent on hypersecretions from the thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other diseases dependent on anterior pituitary hypersecretion; hyperandrogenism, chronic anovulation and other diseases associated with polycystic ovarian syndrome, allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eosinophilia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, glomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders, mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon, myasthenia gravis, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic

autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's disease, Sjogren's syndrome, primary biliary cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vulgarte, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious "anaemia, type 1- diabetes mellitus, cardiovascular conditions and/ or hypertension, and bone conditions such as osteopetrosis and osteoporosis.
In use, the polypeptides of the present invention are typically employed in the form of a pharmaceutical composition in association with a pharmaceutical carrier, diluent and/or excipient, although the exact form of the composition may be tailored to the mode of administration. Administration is preferably to a mammal, more preferably to a human.
The polypeptides may, for example, be employed in the form of an aerosol or nebulisable solution for inhalation or a sterile solution for parenteral administration, intra-articular administration or intra-cranial administration.
For treating endocrine targets, i.v. injection, direct injection into gland, or aerosolisation for lung delivery are preferred; for treating inflammatory cell targets, i.v. injection, sub-cutaneous injection, or surface patch administration or aerosolisation for lung delivery are preferred; for treating exocrine targets, i.v. injection, or direct injection into or direct administration to the gland or aerosolisation for lung delivery are preferred; for treating immunological targets, i.v. injection, or injection into specific tissues eg. thymus, bone marrow, or lymph tissue are preferred; for treatment of cardiovascular targets, i.v. injection is preferred; and for treatment of bone targets, i.v. injection, or direct injection is preferred. In cases of i.v. injection, this should also include the use of pump systems. In the case of compositions for treating neuronal targets, spinal injection (eg. epidural or intrathecal) or indwelling pumps may be used.
The dosage ranges for administration of the polypeptides of the present invention are those to produce the desired therapeutic effect. It will be

appreciated that the dosage range required depends on the precise nature of the components, the route of administration, the nature of the formulation, the age of the patient, the nature, extent or severity of the patient's condition, contraindications, if any, and the judgement of the attending physician.
Suitable daily dosages-are in-the-range-0.0004-1 mg/kg, preferably 0.0001-0.5mg/kg, more preferably 0.002-0.5mg/kg, and particularly preferably 0.004-0.5mg/kg. The unit dosage can vary from less that 1 microgram to 30mg, but typically will be in the region of 0.01 to 1mg per dose, which may be administered daily or preferably less frequently, such as weekly or six monthly.
A particularly preferred dosing regimen is based on 2.5ng of fusion protein as the 1X dose per kg patient. In this regard, preferred dosages are in the range 1X-100X (ie, 2.5 - 250ng). This dosage range is significantly lower (ie. at least 10-fold, typically 100-fold lower) than would be employed with other types "of therapeutic molecules: Moreover, the above-mentioned difference is significantly magnified when the same comparison is made on a molar basis -this is because the fusion proteins of the present invention have a considerably greater molecular weight than the conventional 'small' molecule therapeutics.
Wide variations in the required dosage, however, are to be expected depending on the precise nature of the components, and the differing efficiencies of various routes of administration. For example, oral administration would be expected to require higher dosages than administration by intravenous injection.
Variations in these dosage levels can be adjusted using standard empirical routines for optimisation, as is well understood in the art.
Compositions suitable for injection may be in the form of solutions, suspensions or emulsions, or dry powders which are dissolved or suspended in a suitable vehicle prior to use.

Fluid unit dosage forms are typically prepared utilising a pyrogen-free sterile vehicle. The active ingredients, depending on the vehicle and concentration used, can be either dissolved or suspended in the vehicle.
-Solutions may be-used-for-all-forms of-parenteral administration, and are particularly used for intravenous injection. In preparing solutions the components can be dissolved in the vehicle, the solution being made isotonlc if necessary by addition of sodium chloride and sterilised by filtration through a sterile filter using aseptic techniques before filling into suitable sterile vials or ampoules and sealing. Alternatively, If solution stability is adequate, the solution in its sealed containers may be sterilised by autoclaving.
Advantageously additives such as buffering, solubilising, stabilising, preservative or bactericidal, suspending or emulsifying agents and/or local anaesthetic agents may be dissolved in the vehicle.
Dry powders which are dissolved or suspended in a suitable vehicle prior to use may be prepared by filling pre-sterilised drug substance and other ingredients into a sterile container using aseptic technique in a sterile area.
Alternatively the components (ie. agent plus inhibitor) and other ingredients may be dissolved in an aqueous vehicle, the solution is sterilized by filtration and distributed into suitable containers using aseptic technique in a sterile area. The product is then freeze dried and the containers are sealed asepticafly.
Parenteral suspensions, suitable for intramuscular, subcutaneous or intradermal injection, are prepared in substantially the same manner, except that the sterile components are suspended in the sterile vehicle, instead of being dissolved and sterilisation cannot be accomplished by filtration. The components may be isolated in a sterile state or alternatively it may be sterilised after isolation, eg. by gamma irradiation.

Advantageously, a suspending agent for example polyvinylpyrrolidone is included in the composition/s to facilitate uniform distribution of the components.
Compositions suitable for administration via the respiratory tract include aerosolsrnebulisable-soltitions-or--microfine- powders-for insufflation. In the latter case, particle size of less than 50 microns, especially less than 10 microns, is preferred. Such compositions may be made up in a conventional manner and employed in conjunction with conventional administration devices.
Definitions Section
Targeting Moiety (TM) means any chemical structure associated with an agent that functionally interacts with a Binding Site to cause a physical association between the agent and the surface of a target cell. In the context of the present invention, the target cell is any cell except a nociceptive sensory afferent. The term TM embraces any molecule (ie. a naturally occurring molecule, or a chemically/physically modified variant thereof) that is capable of binding to a Binding Site on the target cell, which Binding Site is capable of internalisation (eg. endosome formation) - also referred to as receptor-mediated endocytosis. The TM may possess an endosomal membrane translocation function, in which case separate TM and Translocation Domain components need not be present in an agent of the present invention.
The TM of the present invention binds (preferably specifically binds) to a target cell.
The term non-cytotoxic means that the protease molecule in question does not kill the target cell to which it has been re-targeted.

The protease of the present invention embraces ail naturally-occurring non-cytotoxic proteases that are capable of cleaving one or more proteins of the exocytic fusion apparatus in eukaryotic cells.
The protease of the present invention is preferably a bacterial protease (or -fragment thereof). -More preferably-the-bacteral-protease-is-selected from the genera Clostridium or Neisseria (eg. a clostridia! L-chain, or a neisserial IgA protease preferably from N. gononhoeao).
The present Invention also embraces modified non-cytotoxic proteases, which include amino acid sequences that do not occur in nature and/or synthetic amino acid residues, so long as the modified proteases still demonstrate the above-mentioned protease activity.
The protease of the present invention preferably demonstrates a serine or metalloprotease activity (eg. endopeptidase activity). The protease is preferably specific for a SNARE protein (eg. SNAP-25, synaptobrevinA/AMP, or syntaxin).
Particular mention is made to the protease domains of neurotoxins, for example the protease domains of bacterial neurotoxins. Thus, the present invention embraces the use of neurotoxin domains, which occur in nature, as well as recombinantly prepared versions of said naturally-occurring neurotoxins.
Exemplary neurotoxins are produced by clostridia, and the term clostridial neurotoxin embraces neurotoxins produced by C. tetani (TeNT), and by C. botulinum (BoNT) serotypes A-G, as well as the closely related BoNT-like neurotoxins produced by C. baratii and C. butyricum. The above-mentioned abbreviations are used throughout the present specification. For example, the nomenclature BoNT/A denotes the source of neurotoxin as BoNT (serotype A). Corresponding nomenclature applies to other BoNT serotypes.

The term L-chain fragment means a component of the L-chain of a neurotdxin, which fragment demonstrates a metalloprotease activity and is capable of proteolytically cleaving a vesicle and/ or plasma membrane associated protein involved In cellular exocytosis.
-A-Translocation-Domain is a molecule that enables-translocation of a protease (or fragment thereof) into a target cell such that a functional expression of protease activity occurs within the cytosol of the target cell. Whether any molecule (eg. a protein or peptide) possesses the requisite translocation function of the present invention may be confirmed by any one of a number of conventional assays.
For example, Shone C. (1987) describes an in vitro assay employing liposomes, which are challenged with a test molecule. Presence of the requisite translocation function is confirmed by release from the liposomes of K+ and/or labelled NAD, which may be readily monitored [see Shone C. (1987) Eur. J. Biochem; vol: 167(1): pp. 175-180].
A further example is provided by Blaustein R. (1987), which describes a simple in vitro assay employing planar phospholipid bilayer membranes. The membranes are challenged with a test molecule and the requisite translocation function is confirmed by an increase in conductance across said membranes [see Blaustein (1987) FEBS Letts; vol. 226, no. 1: pp. 115-120].
Additional methodology to enable assessment of membrane fusion and thus identification of Translocation Domains suitable for use in the present invention are provided by Methods in Enzymology Vol 220 and 221, Membrane Fusion Techniques, Parts A and B, Academic Press 1993.
The Translocation Domain is preferably capable of formation of ion-permeable pores in lipid membranes under conditions of low pH. Preferably it has been found to use only those portions of the protein molecule capable of pore-formation within the endosomal membrane.

The Translocation Domain may be obtained from a microbial protein source, in particular from a bacterial or viral protein source. Hence, in one embodiment, the Translocation Domain is a translocating domain of an enzyme, such as a bacterial toxin or viral protein.
lt is well-documented-that-certain-domains-of-bacterial toxin-molecules are capable of forming such pores. It is also known that certain translocation domains of virally expressed membrane fusion proteins are capable of forming such pores. Such domains may be employed in the present invention.
The Translocation Domain may be of a clostridial origin, namely the HN domain (or a functional component thereof). HN means a portion or fragment of the H-chain of a clostridial neurotoxin approximately equivalent to the amino-terminal half of the H-chain, or the domain corresponding to that fragment in the intact H-chain. It is preferred that the H-chain substantially lacKs the natural binding function of the Hc component of the H-chain. In this regard, the He function may be removed by deletion of the Hc amino acid sequence (either at the DNA synthesis level, or at the post-synthesis level by nuclease or protease treatment). Alternatively, the Hc function may be inactivated by chemical or biological treatment. Thus, the H-chain is preferably incapable of binding to the Binding Site on a target cell to which native clostridial neurotoxin (ie. holotoxin) binds.
In one embodiment, the translocation domain is a HN domain (or a fragment thereof) of a clostridial neurotoxin. Examples of suitable clostridial Translocation Domains include:-
Botulinum type A neurotoxin - amino acid residues (449-871)
Botulinum type B neurotoxin - amino acid residues (441-858)
Botulinum type C neurotoxin - amino acid residues (442-866)
Botulinum type D neurotoxin - amino acid residues (446-862)
Botulinum type E neurotoxin - amino acid residues (423-845)
Botulinum type F neurotoxin - amino acid residues (440-864)

Botulinum type G neurotoxin - amino acid residues (442-863)
Tetanus neurotoxin - amino acid residues (458-879)
For further details on the genetic basis of toxin production in Clostridium botulinum and C. tetani, we refer to Henderson et al (1997) in The Clostridia: Molecular Biology and-Pathogenesis, Academic-press:-
The term HN embraces naturally-occurring neurotoxin HN portions, and modified HN portions having amino acid sequences that do not occur in nature and/or synthetic amino acid residues, so long as the modified HN portions still demonstrate the above-mentioned translocation function.
Alternatively, the Translocation Domain may be of a non-clostridial origin (see Table 1), Examples of non-clostridial Translocation Domain origins include, but not be restricted to, the translocation domain of diphtheria toxin [O=Keefe et al., Proc. Natl. Acad. Sci, USA (1992) 89, 6202-6206; Silverman et a/., J. Biol. Chem. (1993) 269, 22524-22532; and London, E. (1992) Biochem. Biophys. Acta., 1112, pp.25-51], the translocation domain of Pseudomonas exotoxin type A [Prior et al. Biochemistry (1992) 31, 3555-3559], the translocation domains of anthrax toxin [Blanke et al. Proc. Natl. Acad. Sci. USA (1996) 93, 8437-8442], a variety of fusogenic or hydrophobic peptides of translocating function [Plank et al. J. Biol. Chem. (1994) 269, 12918-12924; and Wagner et al (1992) PNAS, 89, pp.7934-7938], and amphiphilic peptides [Murata et al (1992) Biochem., 31, pp. 1986-1992]. The Translocation Domain may mirror the Transiocation Domain present in a naturally-occurring protein, or may include amino acid variations so long as the variations do not destroy the translocating ability of the Translocation Domain.
Particular examples of viral Translocation Domains suitable for use in the present invention include certain translocating domains of virally expressed membrane fusion proteins. For example, Wagner et a/. (1992) and Murata et a/. (1992) describe the translocation (ie. membrane fusion and vesiculation) function of a number of fusogenic and amphiphilic peptides derived from the N-terminal region of influenza virus haemagglutinin. Other virally expressed

membrane fusion proteins known to have the desired translocating activity are a translocating domain of a fusogenic peptide of Semliki Forest Virus (SFV), a translocating domain of vesicular stomatitis virus (VSV) glycoprotein G, a translocating domain of SER virus F protein and a translocating domain of Foamy virus envelope glycoprotein. Virally encoded Aspike proteins have particular-application-in-the-context-of-the-present invention for-example, the E1 protein of SFV and the G protein of the G protein of VSV.
Use of the Translocation Domains listed in Table 1 includes use of sequence variants thereof. A variant may comprise one or more conservative nucleic acid substitutions and/ or nucleic acid deletions or insertions, with the proviso that the variant possesses the requisite translocating function. A variant may also comprise one or more amino acid substitutions and/ or amino acid deletions or insertions, so long as the variant possesses the requisite translocating function.
Table 1

(Table Removed)

SEQ ID NOs
SEQ ID1 DMA sequence of the LC/A
SEQ ID2 DNA sequence of the HN/A
SEQ ID3 DNA sequence of the LC/B
SEQ ID4 DNA sequence of the HN/B
SEQ-ID5 -DNA sequenee-of-the LC/C-
SEQ ID6 DNA sequence of the HN/C
SEQ ID7 DNA sequence of the CP PAR1-B linker
SEQ ID8 DNA sequence of the CP PTH-C linker
SEQ ID9 DNA sequence of the CP PAR1-B fusion
SEQ ID10 Protein sequence of the CP PAR1-B fusion
SEQ ID11 DNA sequence of the CP PTH-C fusion
SEQ 1012 Protein sequence of the CP PTH-C fusion
SEQ ID13 DNA sequence of the CP RGD-C linker
SEQ 1D14 DNA sequence of the CP RGD-C fusion
SEQ ID15 Protein sequence of the CP RGD-C fusion
SEQ ID16 DNA sequence of the CP cyclicRGD-C linker
SEQ ID17 DNA sequence of the CP cyclicRGD-C fusion
SEQ ID18 Protein sequence of the CP cyclicRGD-C fusion
SEQ ID19 DNA sequence of the CP THALWHT-C linker
SEQ 1D20 DNA sequence of the CP THALWHT-C fusion
SEQ ID21 Protein sequence of the CP THALWHT-C fusion
SEQ ID22 DNA sequence of the CP cyclicTHALWHT-C linker
SEQ ID23 DNA sequence of the CP cyclicTHALWHT-C fusion
SEQ ID24 Protein sequence of the CP cyclicTHALWHT-C fusion
SEQ ID25 DNA sequence of the CP ANP-C linker
SEQ ID26 DNA sequence of the CP ANP-C fusion
SEQ ID27 Protein sequence of the CP ANP-C fusion
SEQ ID28 DNA sequence of the CP VIP-C linker
SEQ ID29 DNA sequence of the CP VIP-C fusion
SEQ ID30 Protein sequence of the CP VIP-C fusion
SEQ ID31 DNA sequence of the CP Gastrin releasing peptide -C linker
SEQ ID32 DNA sequence of the CP Gastrin releasing peptide-C fusion
SEQ ID33 Protein sequence of the CP Gastrin releasing peptide -C fusion

EXAMPLES
Example 1 - Preparation of LC/B and HN/B backbone clones
The following procedure creates the LC and HN fragments for use as the component backbone for multidomain fusion expression. This example is based on preparation of a serotype B based clone (SEQ ID3 and SEQ ID4), though the procedures and methods are equally applicable to the other serotypes (illustrated by the sequence listing for serotype A (SEQ ID1 and SEQ ID2) and serotype C (SEQ ID5 and SEQ ID6)).
Preparation of cloning and expression vectors
pCR 4 (Invitrogen) is the chosen standard cloning vector chosen due to the lack of restriction sequences within the vector and adjacent sequencing primer sites for easy construct confirmation. The expression vector is based on the pMAL (NEB) expression vector, which has the desired restriction sequences within the multiple cloning site in the correct orientation for construct insertion (BamHI-Sall-Pstl-Hindlll). A fragment of the expression vector has been removed to create a non-mobilisable plasmid and a variety of different fusion tags have been inserted to increase purification options.
Preparation of protease (eg. LC/B) insert
The LC/B (SEQ ID3) is created by one of two ways:
The DNA sequence is designed by back translation of the LC/B amino acid sequence (obtained from freely available database sources such as GenBank (accession number P10844) or Swissprot (accession locus BXB_CLOBO) using one of a variety of reverse translation software tools (for example EditSeq best E coli reverse translation (DNASTAR Inc.), or Backtranslation tool v2.0 (Entelechon)). BamHI/Sa/l recognition sequences are incorporated at the 5' and 3' ends respectively of the sequence maintaining the correct reading frame. The DNA sequence is screened (using software such as

MapDraw, DNASTAR Inc.) for restriction enzyme cleavage sequences incorporated during the back translation. Any cleavage sequences that are found to be common to those required by the cloning system are removed manually from the proposed coding sequence ensuring common E. coli codon usage is maintained. E. coli codon usage is assessed by reference to software programs such as Graphical Codon Usage Analyser (Geneart), and the %GC content and codon usage ratio assessed by reference to published codon usage tables (for example GenBank Release 143, September 13 2004). This optimised DNA sequence containing the LC/B open reading frame (ORF) is then commercially synthesized (for example by Entelechon, Geneart or Sigma-Genosys) and is provided in the pCR 4 vector.
The alternative method is to use PCR amplification from an existing DNA sequence with BamHI and Sall restriction enzyme sequences incorporated into the 5' and 3' PCR primers respectively. Complementary oligonucleotide primers are chemically synthesised by a Supplier (for example MWG or Sigma-Genosys) so that each pair has the ability to hybridize to the opposite strands (3' ends pointing "towards" each other) flanking the stretch of Clostridium target DNA, one oligonucleotide for each of the two DNA strands. To generate a PCR product the pair of short oligonucleotide primers specific for the Clostridium DNA sequence are mixed with the Clostridium DNA template and other reaction components and placed in a machine (the 'PCR machine') that can change the incubation temperature of the reaction tube automatically, cycling between approximately 94 °C (for denaturation), 55 °C (for oligonucleotide annealing), and 72 °C (for synthesis). Other reagents required for amplification of a PCR product include a DNA polymerase (such as Tag or Pfu polymerase), each of the four nucleotide dNTP building blocks of DNA in equimolar amounts (50-200 µM) and a buffer appropriate for the enzyme optimised for Mg2+ concentration (0.5-5 mM).
The amplification product is cloned into pCR 4 using either, TOPO TA cloning for Tag PCR products or Zero Blunt TOPO cloning for Pfu PCR products (both

kits commercially available from Invitrogen). The resultant clone is checked by sequencing. Any additional restriction sequences that are not compatible with the cloning system are then removed using site directed mutagenesis (for example using Quickchange (Stratagene Inc.)).
Preparation of translocation (eg. HN) insert
The HN/B (SEQ ID4) is created by one of two ways:
The DNA sequence is designed by back translation of the HN/B amino acid sequence (obtained from freely available database sources such as GenBank (accession number P10844) or Swissprot (accession locus BXB_CLOBO)) using one of a variety of reverse translation software tools (for example EditSeq best E. coli reverse translation (DNASTAR Inc.), or Backtranslation tool v2.0 (Entelechon)). A Pstl restriction sequence added to the N-terminus and Xbal-stop codon-Hindlll to the C-terminus ensuring the correct reading frame in maintained. The DNA sequence is screened (using software such as MapDraw, DNASTAR Inc.) for restriction enzyme cleavage sequences incorporated during the back translation. Any sequences that are found to be common to those required by the cloning system are removed manually from the proposed coding sequence ensuring common E. coli codon usage is maintained. E. coli codon usage is assessed by reference to software programs such as Graphical Codon Usage Analyser (Geneart), and the %GC content and codon usage ratio assessed by reference to published codon usage tables (for example GenBank Release 143, September 13 2004). This optimised DNA sequence is then commercially synthesized (for example by Entelechon, Geneart or Sigma-Genosys) and is provided in the pCR 4 vector.
The alternative method is to use PCR amplification from an existing DNA sequence with Pstl and Xbal-stop codon-H/ndlll restriction enzyme sequences incorporated into the 5' and 3' PCR primers respectively. The PCR amplification is performed as described above. The PCR product is inserted into pCR 4 vector and checked by sequencing. Any additional restriction sequences that are not compatible with the cloning system are then removed

using site directed mutagenesis (for example using Quickchange (Stratagene Inc.)).
Example 2 - Preparation of a LC/B-PAR1-HN/B fusion protein
Preparation of linker-PAR1-spacer insert
The L.C-HN linker can be designed from first principle, using the existing sequence information for the linker as the template. For example, the serotype B linker defined as the inter-domain polypeptide region that exists between the cysteines of the disulphide bridge between LC and HN within which proteolytic activation occurs. This sequence Information is freely available from available database sources such as GenBank (accession number P10844) or Swissprot (accession locus BXB_CLOBO). It is into this linker that an Enterokinase site, PAR1 and spacer are incorporated and using one of a variety of reverse translation software tools (for example EditSeq best E. co/i reverse translation (DNASTAR Inc.), or Backtranslation tool v2.0 (Entelecnon)), the DNA sequence encoding the linker-ligand-spacer region is determined. Restriction site are then incorporated into the DNA sequence and can be arranged as BamHI-Sa/l-linker-protease site-PAR1-Mhel-spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID7). It is important to ensure the correct reading frame is maintained for the spacer, PAR1 and restriction sequences and that the Xbal sequence is not preceded by the bases, TC which would result on DAM methylation. The DNA sequence is screened for restriction sequence incorporated and any additional sequences are removed manually from the remaining sequence ensuring common E. coli codon usage is maintained. E. coli codon usage is assessed by reference to software programs such as Graphical Codon Usage Analyser (Geneart), and the %GC content and codon usage ratio assessed by reference to published codon usage tables (for example GenBank Release 143, September 13 2004). This optimised DNA sequence is then commercially synthesized (for example by Entelechon, Geneart or Sigma-Genosys) and is provided in the pCR 4 vector.

Preparation of the LC/B-PAR1-HN/B fusion
In order to create the LC-linker-PAR1-spacer-HN construct (SEQ ID9), the pCR 4 vector encoding the linker (SEQ ID7) is cleaved with BamHI + Sa/l restriction enzymes. This cleaved vector then serves as the recipient vector for insertion and ligation of the LC/B DNA (SEQ IDS) cleaved with BamHI + Sa/l. The resulting plasmid DNA is then cleaved with Pstl + Xbal restriction enzymes and serves as the recipient vector for the insertion and ligation of the HN/B DNA (SEQ ID4) cleaved with Pstl + Xbal. The final construct contains the LC-linker-PAR1-spacer-HN ORF (SEQ ID9) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQID10.
Example 3 - Preparation LC/C-PTH-HN/C fusion protein
The LC-HN linker can be designed using the methods described in example two but using the C serotype linker arranged as BamHI-Sa/I-linker-protease site-PTH-Nhel-spacer-Spel-Pstl-Xbal-stop codon-H//7dlII (SEQ ID8). The LC/C-PTH-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ 106) made using the methods described in example one and constructed using methods described in example two. The final construct contains the LC-linker-PTH-spacer-HN ORF (SEQ ID 1i) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID 12.
Example 4 - Preparation and purification of LC/C-RGD-HN/C fusion protein
The LC-HN linker is designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sa/l-linker-protease site-RGD-Nhel-spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID 13). The LC/C-RGD-HN/C fusion is then assembled using the LC/C (SEQ ID 5) and HN/C (SEQ ID 6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC-linker-

RGD-spacer-HN ORF (SEQ ID 14) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID15, The resultant expression plasmid, pMAL LC/C-RGD-HN/C is transformed into E coli BL21 for recombinant protein expression.
Expression of LC/C-RGP-HN/C fusion protein
Expression of the LC/C-RGD-HN/C fusion protein is achieved using the following protocol. Inoculate 100ml of modified TB containing 0.2% glucose and 100µg/ml ampicillin in a 250ml flask with a single colony from the LC/C-RGD-HN/C expression strain. Grow the culture at 37°C, 225 rpm for 16 hours. Inoculate 1L of modified TB containing 0.2% glucose and 100µg/ml ampicillin !n a 2L flask with 10ml of overnight culture. Grow cultures at 37°C until an approximate OD600nm of 0.5 is reached at which point reduce the temperature to 16°C. After 1 hour induce the cultures with 1mM IPTG and grow at 16°C for a further-16-hours. -Figure 1 demonstrates the expressed protein in E. coli as analysed by SDS-PAGE,
Purification of LC/C-RGD-HN/C fusion protein
Defrost falcon tube containing 25ml 50mM HEPES pH 7.2 200mM NaCI and approximately 10 g of E. coli BL21 cell paste. Sonicate the cell paste on ice 30 seconds on, 30 seconds off for 10 cycles at a power of 22 microns ensuring the sample remains cool. Spin the lysed cells at 18 000 rpm, 4°C for 30 minutes. Load the supernatant onto a 0.1 M NiSO4 charged Chelating column (20-30 ml column is sufficient) equilibrated with 50mM HEPES pH 7.2 200mM NaCI. Using a step gradient of 10 and 40mM imidazole, wash away the non-specific bound protein and elute the fusion protein with 100mM imidazole. Dialyse the eluted fusion protein against 5L of 50mM HEPES pH 7,2 200mM NaCI at 4°C overnight and measure the OD of the dialysed fusion protein. Add 1 unit of factor Xa per 100 µg fusion protein and incubate at 25°C static overnight. Load onto a 0.1 M NiSO4 charged Chelating column (20-30 ml column is sufficient) equilibrated with 50 mM HEPES pH 7.2 200mM

Wash column to baseline with 50mM HEPES pH 7.2 200mM NaCl. Using a step gradient of 10 and 40mM imidazole, wash away the non-specific bound protein and elute the fusion protein with 100mM imidazole. Dialyse the eluted fusion protein against 5L of 50mM HEPES pH 7.2 200mM NaCl at 4°C overnight and concentrate the fusion to about 2 mg/ml, aliquot sample and freeze at -20°C. Test purified protein using OD, BCA and purity analysis. Figure 2 demonstrates the purified protein as analysed by SDS-PAGE.
Example 6 - Preparation LC/C-cyclicRGD-HN/C fusion protein
The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sa/l-linker-protease site-cylicRGD-Nhel~spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID16) The LC/C-cyclicRGD-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC-linker-cyclicRGD-spacer-HN ORF (SEQ 1017) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID18. The resultant expression plasmid, pMAL LC/C-cyclicRGD-HN/C was transformed into E. coli BL21 for recombinant protein expression. Expression of the fusion protein was carried out as described in Example 4. Figure 1 demonstrates the protein expressed in E. coli as analysed by SDS-PAGE.
Example 6 - Preparation LC/C-THALWHT-HN/C fusion protein
The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype (inker arranged as BamHI-Sall-linker-protease site-THALWHT-Nhel-spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID19). The LC/C-THALWHT-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC~linker-THALWHT-spacer-HN ORF (SEQ ID20) for transfer into

expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID21. Expression of the fusion protein was carried out as described in Example 4. Figure 1 demonstrates the protein expressed in E. coli as analysed by SDS-PAGE.
The-THALWHT peptide sequence given in this Example (SEQ IDs 19, 20 and 21) can be exchanged with another peptide sequence found by phage display techniques. For example, LEBP-1 (QPFMQCLCLIYDASC), LEBP-2 (RNVPP1FNDVYWIAF) and LEBP-3 (VFRVRPWYQSTSQS) (Wu ef a/., 2003); CDSAFVTVDWGRSMSLC (Florea ef a/., 2003); SERSMNF, YGLPHKF, PSGAARA, LPHKSMP, LQHKSMP (Writer ef a/., 2004); FSLSKPP, HSMQLST and STQAMFQ peptides (Rahim ef a/., 2003).
Example 7 - Preparation LC/C-cyclicTHALWHT-HN/C fusion protein
The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sall-linker-protease site-cyclicTHALWHT-Nhel-spacer-Spel-Pstl-X6al-stop codon-Hindlll (SEQ ID22). The LC/C-cyciicTHALWHT-Hn/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in example one and constructed using methods described in Example 2. The final construct contains the LC-linker-cyclicTHALWHT-spacer-HN ORF (SEQ ID23) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ 1D24. Expression of the fusion protein was carried out as described in Example 4. Figure 1 demonstrates the protein expressed in E. coli as analysed by SDS-PAGE.
The THALWHT peptide sequence given in this Example (SEQ IDs 19, 20 and 21) can be exchanged with another peptide sequence found by phage display techniques. For example, LEBP-1 (QPFMQCLCLIYDASC), LEBP-2 (RNVPPIFNDVYWIAF) and LEBP-3 (VFRVRPWYQSTSQS) (Wu ef a/., 2003); CDSAFVTVDWGRSMSLC (Florea ef a/., 2003); SERSMNF, YGLPHKF, PSGAARA, LPHKSMP, LQHKSMP (Writer ef a/., 2004); FSLSKPP, HSMQLST and STQAMFQ peptides (Rahim ef a/., 2003).

Example 8 ~ Preparation LC/C-ANP-HN/C fusion protein
The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sa/l-linker-protease site-ANP-Nhel-spacer-Spel-Pstl-Xbal-stop eodon-Hindlll (SEQ ID25). The LC/C-ANP-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC-linker-ANP-spacer-HN ORF (SEQ ID26) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID27.
Example 9 - Preparation LC/C-VIP-HN/C fusion protein
The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sall-linker-protease site-VIP-Nhel-spacer-Spel-Pstl-Xbal-stop codon-Hindlll (SEQ ID28). The LC/C-VIP-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods described in Example 1 and constructed using methods described in Example 2. The final construct contains the LC-linker-VIP-spacer-HN ORF (SEQ ID29) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID30.
The VIP sequence given in SEQ ID28 could be replaced with VIP analogue or agonist sequences. For example, [R15'20'21, L17]-VIP or [R15-20-21, L17]-VIP-GRR (Kashimoto et al,, 1996; Onoue et a/., 2004), [A2,8'9'16l19'24]-VIP or
Example 10 - Preparation LC/C-gastrin releasing peptide-HN/C fusion protein

The LC-HN linker can be designed using the methods described in Example 2 but using the C serotype linker arranged as BamHI-Sa/l-linker-protease site-gastrin releasing peptide-Nhel-spacer-Spel-Psfl-X6al-stop codon-Hindlll (SEQ ID34). The LC/C-gastrin releasing peptide-HN/C fusion is then assembled using the LC/C (SEQ ID5) and HN/C (SEQ ID6) made using the methods-described-in Example 1-and-construeted-using methods described in Example 2. The final construct contains the LC-linker-gastrin releasing peptide-spacer-HN ORF (SEQ ID35) for transfer into expression vectors for expression to result in a fusion protein of the sequence illustrated in SEQ ID36.
Example 11 - Assessment of functionality of the LC/C-RGD-HN/C fusion protein
The functionality of the TM component of the LC/C-RGD-HN/C fusion protein (prepared according to Example 4) is assessed by a ligand binding assay. To facilitate assessment of ligand binding, an RGD binding-peptide is synthesised in a biotinylated and non-biotinylated form. Binding of the fusion protein is determined by a competition assay with the biotinylated form. Briefly, NCI-H292 cells are plated into 96 well plates and viable cultures established. Cells and solutions are are-chilled to 4°C and solutions are prepared in cell feeding medium-plus-HEPES (50mM). Prior to treatment, media is removed from the cells and replaced with media-plus-HEPES (500µI per well), which is then also removed. Labelled ligand, at x 2 the required concentration, is added to all wells (50µl per well). The fusion protein, at x 2 the required concentration, is then added to wells (50µl per well). After 1 hour at 4°C, the media is removed and replaced with media + HEPES (100µl per well). This media is removed and replaced with media + HEPES (100µl per well). Cells are lysed with 100µl per well PBS-Tween 0.1% for 5 mins at 4°C. PBS-Tween is removed and cells are washed with media + HEPES (100µl per well). This media is removed and replaced with 100µl PBS + 100µl streptavidin-HRP per well. Cells are incubated at RTP for 20 mins. The PBS + streptavidin is removed and the cells are washed with PBS-Tween. 100µl per well of TMB is added and the cells are incubated at 37°C for 10mins. 50µl per well 2M H2SO4 is

and the plate read at 450nm. Using this methodology, the ability of the TM component of the LC/C-RGD-HN/C fusion protein to bind to the cell surface is confirmed.
Description of the Figures:
Figure 1 • Expression of LC/C-RGD-HN/C, LC/C-cyclicRGD-HN/C, LC/C-THALWHT-HN/C and LC/C-cycllcTHALWHT-HN/C fusion proteins in E. coll.
Using the methodology outlined in Example 4, LC/C-RGD-HN/C, LC/C-cyclic RGD-HN/C, LC/C-THALWHT-HN/C and LC/C-cyclicTHALWHT-HN/C fusion proteins were expressed in E. coli BL21 cells. Briefly, 1L of TB media containing 0.2% glucose and 100pg/ml ampicillin was inoculated with 10ml of starter culture. Cultures were grown at 37°C until an approximate OD600nm of 0.5 was reached at which point the temperature was reduced to 16°C. After 1 hour the cultures were induced with 1mM IPTG and grown for a further 16 hours.
Lane 1, LC/C-THALWHT-HN/C;
Lane 3, LC/C-cyclicTHALWHT-HN/C; Lane 4t LC/C-cyclicRGD-HN/C.
Figure 2 - Purification of a LC/C-RGD-HN/C fusion protein
Using the methodology outlined in Example 5, a LC/C-RGD-HN/C fusion protein was purified from E. coli BL21 cells. Briefly, the soluble products obtained following cell disruption were applied to a nickel-charged affinity capture column. Bound proteins were eluted with 100mM imidazole, treated with Factor Xa to activate the fusion protein and remove the maltose-binding protein (MBP) tag, then re-applied to a second nickel-charged affinity capture column. Samples from the purification procedure were assessed by SDS-PAGE. The final purified material in the absence and presence of reducing agent is identified in the lanes marked [-] and [+] respectively.

References
Florea etal., (2003) J. Drug Targeting 11: 383-390 Jost et al., (2001) FEBS lett. 489: 263-269 Lee et al., (2001) Eur. J. Biochem. 268: 2004-2012 Mathias et al., (1994) J. Virol. 68: 6811-6814 Rahim et al., (2003) Biotechniques 35: 317-324 Roivaninen et a/., (1991) J. ViroL 65:4735-4740 Ruoslahti (1996) Ann. Rev. Ceil Dev. Biol. 12: 697-715 Schneider etal., (1999) FEBS left. 458: 329-332 Writer et a/., (2004) J. Drug Targeting 12:185-193 Wu et a/., (2003) Gene Ther. 10:1429-1436

We Claims
1. A single chain, polypeptide fusion protein, comprising:
(a) a non-cytotoxic protease, or a fragment thereof, which protease or protease fragment is capable of cleaving a protein of the exocytic fusion apparatus of a target cell;
(b) a Targeting Moiety that is capable of binding to a Binding Site on the target cell, which Binding Site is capable of undergoing endocytosis to be incorporated into an endosome within the target cell;
(c) a protease cleavage site at which site the fusion protein is cleavable by a protease, wherein the protease cleavage site is located between the non-cytotoxic protease or fragment thereof and the Targeting Moiety; and
(d) a translocation domain that is capable of translocating the protease or protease fragment from within an endosome, across the endosomal membrane and into the cytosol of the target cell;
wherein the Targeting Moiety is located between the protease cleavage site and the translocation domain.
2. The fusion protein according to Claim 1, wherein the Targeting Moiety and the protease cleavage site are separated by at most 10 amino acid residues, preferably by at most 5 amino acid residues, and more preferably by at most zero amino acid residues.
3. The fusion protein according to any preceding claim, wherein the non-cytotoxic protease is a clostridial neurotoxin L-chain.
4. The fusion protein according to any preceding claim, wherein the translocation domain is the HN domain of a clostridial neurotoxin.
5. The fusion protein according to any preceding claim, wherein the Targeting Moiety comprises at most 50 amino acid residues, preferably at

most 40 amino acid residues, more preferably at least 30 amino acid residues, and most preferably at most 20 amino acid residues.
6. The fusion protein according to any preceding claim, wherein the Targeting Moiety comprises a PAR ligand, preferably the ligand PARI.
7. The fusion protein according to any of Claims 1-5, wherein the Targeting Moiety comprises a ligand that binds to PTH-1, preferably the ligand comprises a PTH peptide; or wherein the Targeting Moiety comprises a linear or cyclic integrin binding sequence, preferably a triple Arg-Gly-Asp (RGD) binding sequence.
8. The fusion protein according to any preceding claim, wherein the Targeting Moiety binds to a cell selected from the group consisting of: a mucus-secreting cell or a neuronal cell controlling or directing mucus secretion; an endocrine cell; an inflammatory cell; an exocrine cell; an immunological cell; a cardiovascular cell; or a bone cell.
9. The fusion protein according to any of Claims 1-8 wherein the Targeting Moiety comprises a ligand selected from the group consisting of: TFLLR ; PAR-1 ; PTH; VIP or a VIP analogue selected from the group consisting of [R15,20,21, L17] - VIP, [R15'20'21, L17] - VIP-GRR, [A2'8'9'16'19'24] - VIP or [A2'8'9'16'19'24'25] -VIP ; beta2 adrenoreceptor agonists; gastrin-releasing peptide (GRP); calcitonin gene related peptide (CGRP); thyroid-stimulating hormone (TSH); insulin; insulin-like growth factor; TSH-releasing hormone (protirelin); FSH/LH-releasing hormone (gonadorelin); corticotrophin releasing hormone (CRH); adenocorticotrophic hormone (ACTH); pituitary adenyl cyclase activating peptide; a ligand to the C4 domain of the Fc IgE; a ligand to the C3a/C4a-R complement receptor; an antigen reactive towards the CR4 complement receptor; macrophage stimulating factor; an antigen associated with the iC3b complement receptor; IL8; Epstein Barr virus fragment/surface feature ; thrombin; thrombin receptor agonist peptide (TRAP); GP1b surface antigen-recognising antibodies ; calcitonin ; the osteoclast differentiation factor TRANCE, RANKL, or OPGL ; a linear or cyclic peptide selected from the group consisting of: THALWHT, LEBP-1

(QPFMQCLCLIYDASC), LEBP-2 (RNVPPIFNDVYWIAF), LEBP-3
(VFRVRPWYQSTSQS), CDSAFVTVDWGRSMSLC, SERSMNF, YGLPHKF, PSGAARA, LPHKSMP, LQHKSMP, FSLSKPP, HSMQLST, STQAMFQ ; and ANP.
10. The fusion protein according to any preceding claim, wherein the fusion protein comprises a purification tag.
11. The fusion according to Claim 10, wherein the fusion protein comprises a purification tag, which is present at the N-terminal and/ or C-terminal end of the fusion protein.
12. The fusion protein according to Claims 10 or 11, wherein the purification tag is joined to the fusion protein by a peptide spacer molecule.
13. The fusion protein according to any preceding claim, wherein the translocation domain is separated from the Targeting Moiety by a peptide spacer molecule.
14. A polypeptide fusion protein comprising any one of SEQ ID NOs: 10, 12, 15, 18, 21,24, 27, 30, and 33.
15. A nucleic acid sequence encoding the polypeptide fusion protein according to any preceding Claim.
16. A nucleic acid sequence according to Claim 15, wherein the nucleic acid molecule comprises any one of SEQ ID NOs: 1-9, 11, 14, 17, 20, 23, 26, 29, and 32.
17. A DNA vector, which comprises a promoter, a nucleic acid sequence according to Claim 15 or Claim 16, wherein said DNA sequence is located downstream of the promoter, and a terminator is located downstream of the DNA construct.

18. The complementary DNA strand of the DNA sequence according to Claim 15 or Claim 16.
19. A method for preparing a single-chain polypeptide fusion protein according to any of Claims 1-14, comprising expressing a nucleic acid sequence according to Claim 15 or Claim 16, or a DNA vector according to Claim 17, in a host cell.
20. A method of preparing a non-cytotoxic agent, comprising:
(a) contacting a single-chain polypeptide fusion protein according to
any of Claims 1-14 with a protease capable of cleaving the protease
cleavage site;
(b) cleaving the protease cleavage site;
and thereby forming a di-chain fusion protein.
21. A non-cytotoxic polypeptide, obtainable by the method of Claim 20,
wherein the polypeptide is a di-chain polypeptide, and wherein:
(a) the first chain comprises the non-cytotoxic protease, or a fragment thereof, which protease or protease fragment is capable of cleaving a protein of the exocytic fusion apparatus of a target cell;
(b) the second chain comprises the TM and the translocation domain that is capable of translocating the protease or protease fragment from within an endosome, across the endosomal membrane and into the cytosol of the target cell; and
the first and second chains are disulphide linked together.
22. Use of a fusion protein according to any of Claims 1-14 or a
polypeptide according to Claim 21, for the manufacture of a medicament for
treating, preventing or ameliorating a medical condition or disease selected
from the group consisting of mucus hypersecretion, asthma, and/ or chronic
obstructive pulmonary disease, endocrine neoplasia including MEN,
thyrotoxicosis and other diseases dependent on hypersecretions from the
thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other
diseases dependent on anterior pituitary hypersecretion; hyperandrogenism,

chronic anovulation and other diseases associated with polycystic ovarian syndrome, allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eosinophilia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, glomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders, mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon, myasthenia gravis, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's disease, Sjogren's syndrome, primary biliary cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vulgaris, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious anaemia, type 1 diabetes mellitus, cardiovascular conditions and/ or hypertension, and bone conditions such as osteopetrosis and osteoporosis.
23. A method of treating, preventing or ameliorating a medical condition or disease in a subject, comprising administering to said patient a therapeutically effective amount of a fusion protein according to any of Claims 1-14 or a polypeptide according to Claim 21, wherein the medical condition or disease is selected from the group consisting of mucus hypersecretion, asthma, and/ or chronic obstructive pulmonary disease, endocrine neoplasia including MEN, thyrotoxicosis and other diseases dependent on hypersecretions from the thyroid; acromegaly, hyperprolactinaemia, Cushings disease and other diseases dependent on anterior pituitary hypersecretion; hyperandrogenism, chronic anovulation and other diseases associated with polycystic ovarian syndrome, allergies (seasonal allergic rhinitis (hay fever), allergic conjunctivitis, vasomotor rhinitis and food allergy), eosinophilia, asthma, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus

erythematosus, ulcerative colitis, Crohn's disease, haemorrhoids, pruritus, glomerulonephritis, hepatitis, pancreatitis, gastritis, vasculitis, myocarditis, psoriasis, eczema, chronic radiation-induced fibrosis, lung scarring and other fibrotic disorders, mucus hypersecretion from mucus-secreting cells located in the alimentary tract, in particular located in the colon, myasthenia gravis, rheumatoid arthritis, systemic lupus erythematosus, discoid lupus erythematosus, organ transplant, tissue transplant, fluid transplant, Graves disease, thyrotoxicosis, autoimmune diabetes, haemolytic anaemia, thrombocytopenic purpura, neutropenia, chronic autoimmune hepatitis, autoimmune gastritis, pernicious anaemia, Hashimoto's thyroiditis, Addison's disease, Sjogren's syndrome, primary biliary cirrhosis, polymyositis, scleroderma, systemic sclerosis, pemphigus vulgaris, bullous pemphigoid, myocarditis, rheumatic carditis, glomerulonephritis (Goodpasture type), uveitis, orchitis, ulcerative colitis, vasculitis, atrophic gastritis, pernicious anaemia, type 1 diabetes mellitus, cardiovascular conditions and/ or hypertension, and bone conditions such as osteopetrosis and osteoporosis

"FUSION PROTEINS COMPRISING A NON-CYTOTOXIC PROTEASE, A TARGETING MOIETY, A PROTEASE CLEAVAGE SITE AND A TRANSLOCATION DOMAIN"

Documents:

Inventors:

PCT Conventions: